Artificial intelligence is moving beyond software and further into the physical side of business. Companies in food production and logistics are starting to use data systems to support day-to-day decisions, not long-term planning.
That change is visible in The Hershey Company’s latest strategy update. At its Investor Day, the company said it plans to use AI in its operations, from sourcing analytics to plant automation and fulfilment, with a focus on how the business runs behind the scenes.
Hershey said it plans to apply AI to sourcing and fulfilment. This includes using data to guide how ingredients are bought and how products are distributed. In its Investor Day material, the company said it aims to build “a faster, smarter and more resilient supply chain powered by automation and AI-enabled decision making”.
Supply chains in food and snack markets are under steady pressure: Costs can change quickly, demand can change by season, by market, or by product category, and retailers still expect goods to arrive on time and in the right mix.
Hershey said its digital planning tools are meant to connect different parts of the business. The company said those systems are designed to reduce waste and improve inventory levels. It also said digital operational planning can connect data in the supply chain and help raise service levels.
From reporting to action
Part of Hershey’s update is its use of the phrase “AI-enabled decision-making.” The company said its approach will link sourcing and delivery more closely and plans to use automated fulfilment systems for custom assortments and to improve speed to market.
This is a useful way to read strategy. A hard task is turning data into decisions that help operations move faster or with fewer mistakes.
This is where AI is starting to play a ******* role, according to Hershey’s. The value comes from how operations are connected.
AI in the supply chain and plant operations
The changes also extend into manufacturing. Hershey said it will increase plant automation to improve manufacturing efficiency and use AI in more parts of its operating model. What is changing is how AI fits into those systems. Instead of sitting apart from production, it is being positioned as part of the process used to guide planning and support execution.
That may help companies improve planning and respond more quickly when conditions change. In a business where input costs and consumer demand can change often, even small gains in timing can matter.
Food and snack companies deal with constant swings in input costs and demand. Ingredients like cocoa and sugar are affected by weather, trade flows, and supply issues. Companies still have to keep factories running and products moving through retail channels.
Hershey’s plan to use sourcing analytics is one example of how AI may be applied in that setting. By analysing supplier data and market trends, the company may improve how it buys raw materials and manages risk. The company also said it wants to better connect workers in its operations. That suggests the strategy is not only about automation. It is also about coordination in the business.
Hershey said it plans to “incorporate AI in every stage of its operations,” including sourcing analytics and worker connectivity, as well as automated fulfilment and plant automation.
That makes the company a useful case study for a wider change in enterprise AI. Firms are moving away from narrow pilots and toward broader use in business functions. In that model, AI is treated as a part of supply and delivery systems.
CEO Kirk Tanner framed the plan around growth and execution, saying, “The strategy is clear. The team is ready. The next chapter of growth and leading performance starts now”.
Where this may lead
The kind of change is likely to spread as more companies look for ways to connect data with operational decisions. Hershey’s strategy shows how AI is starting to take a larger role in industries built on physical goods. The technology may sit in the background, but its role in daily operations is becoming harder to ignore.
(Photo by Janne Simoes)
See also: JPMorgan begins tracking how employees use AI at work
Want to learn more about AI and big data from industry leaders? Check outAI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Hershey applies AI across its supply chain operations appeared first on AI News.
View the full article
Heavy industry relies on people to inspect hazardous, dirty facilities. It’s expensive, and putting humans in these zones carries obvious safety risks. Swiss robot maker ANYbotics and software company SAP are trying to change that.
ANYbotics’ four-legged autonomous robots will be connected straight into SAP’s backend enterprise resource planning software. Instead of treating a robot as a standalone asset, this turns it into a mobile data-gathering node within an industrial IoT network.
This initiative shows that hardware innovation can now effectively connect with established business workflows. Underscoring that broader trend, SAP is sponsoring this year’s AI & Big Data Expo North America at the San Jose McEnery Convention Center, CA, an event that is fittingly co-located with the IoT Tech Expo and Intelligent Automation & Physical AI Summit.
When equipment breaks at a chemical plant or offshore rig, it costs a fortune. People do routine inspections to catch these issues early, but humans get tired and plants are massive. Robots, on the other hand, can walk the floor constantly, carrying thermal, acoustic, and visual sensors. Hook those sensors into SAP, and a hot pump instantly generates a maintenance request without waiting for a human to report it.
Cutting out the reporting lag
Usually, finding a problem and logging a work order are two disconnected steps. A worker might hear a weird noise in a compressor, write it down, and type it into a computer hours later. By the time the replacement part gets approved, the machine might be wrecked.
Connecting ANYbotics to SAP eliminates that delay. The robot’s onboard AI processes what it sees and hears instantly. If it hears an irregular motor frequency, it doesn’t just flash a warning on a separate screen, it uses APIs to tell the SAP asset management module directly. The system immediately checks for spare parts, figures out the cost of potential downtime, and schedules an engineer.
This automates the flow of information from the floor to management. It also means machinery gets judged on hard, consistent numbers instead of a human inspector’s subjective opinion.
Putting robots in heavy industry isn’t like installing software in an office—companies have to deal with unreliable infrastructure. Factories usually have awful internet connectivity due to thick concrete, metal scaffolding, and electromagnetic interference.
To make this work, the setup relies on edge computing. It takes too much bandwidth to constantly stream high-def thermal video and lidar data to the cloud. So, the robots crunch most of that data locally. Onboard processors figure out the difference between a machine running normally and one that’s dangerously overheating. They only send the crucial details (i.e. the specific fault and its location) back to SAP.
To handle the network issues, many early adopters build private 5G networks. This gives them the coverage they need across huge facilities where regular Wi-Fi fails. It also locks down access, keeping the robot’s data safe from interception.
Of course, security is a major issue. A walking robot packed with cameras is effectively a roaming vulnerability. Companies must use zero-trust network protocols to constantly verify the robot’s identity and limit what SAP modules it can touch. If the robot gets hacked, the system has to cut its connection instantly to stop the attackers from moving laterally into the corporate network.
These robots generate a massive amount of unstructured data as they walk around. Turning raw audio and thermal images into the neat tables SAP requires is difficult.
If companies don’t manage this right, maintenance teams will drown in alerts. A robot that is too sensitive might ***** out hundreds of useless warnings a day, making the SAP dashboard completely ignored. IT teams have to set strict rules before turning the system on. They need exact thresholds for what triggers a real maintenance ticket and what just needs to be watched.
The setup usually uses middleware to translate the robot’s telemetry into SAP’s language. This software acts as a filter, throwing out the noise so only actual problems reach the ERP system. The data lake storing all this information also needs to be organised for future machine learning projects. Fixing broken machines is the short-term goal; the long-term payoff is using years of robot data to predict failures before they happen.
Ensuring a successful physical AI deployment
Dropping robots into a factory naturally makes people nervous. The project’s success often comes down to how human resources handles it. Workers usually look at the robots and assume layoffs are next.
Management has to be clear about why the robots are there. The goal is to get people out of dangerous areas like high-voltage zones or toxic chemical sectors to reduce injuries. The robot collects the data, and the human engineer shifts to analysing that data and doing the actual repairs.
This requires retraining. Workers who used to walk the perimeter now have to read SAP dashboards, manage automated tickets, and work with the robots. They have to trust the sensors, and management has to make sure operators know they can take manual control if something unexpected happens.
Companies need to take the rollout slowly. Because syncing physical robots with enterprise software is complicated, large-scale rollouts should start as small, targeted pilots.
The first test should be in one specific area with known hazards but rock-solid internet. This lets IT watch the data flow between the hardware and SAP in a controlled space. At this stage, the main job is making sure the data matches reality. If the robot sees one thing and SAP records another, it has to be audited and fixed daily.
Once the data pipeline actually works, the company can add more robots and connect other systems, like automated parts ordering. IT chiefs have to keep checking if their private networks can handle more robots, while security teams update their defenses against new threats.
If companies treat these autonomous inspectors as an extension of their corporate data architecture, they get a massive amount of information about their physical assets. But pulling it off means getting the network infrastructure, the data rules, and the human element exactly right.
See also: The rise of invisible IoT in enterprise operations
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post SAP and ANYbotics drive industrial adoption of physical AI appeared first on AI News.
View the full article
Financial institutions are learning to deploy compliant AI solutions for greater revenue growth and market advantage.
For the better part of ten years, financial institutions viewed AI primarily as a mechanism for pure efficiency gains. During that era, quantitative teams programmed systems designed to discover ledger discrepancies or eliminate milliseconds from automated trading execution times. As long as the quarterly balance sheets reflected positive gains, stakeholders outside the core engineering groups rarely scrutinised the actual maths driving these returns.
The arrival of generative applications and highly complex neural networks completely dismantled that widespread state of comfortable ignorance. Today, it’s not acceptable for banking executives to approve new technology rollouts based simply on promises of accurate predictive capabilities.
Across Europe and North America, lawmakers are aggressively drafting legislation aimed at punishing institutions that utilise opaque algorithmic decision-making processes. Consequently, the dialogue within corporate boardrooms has narrowed intensely to focus on safe AI deployment, ethics, model oversight, and legislation specific to the financial industry.
Institutions that choose to ignore this impending regulatory reality actively place their operational licenses in jeopardy. However, treating this transition purely as a compliance exercise ignores the immense commercial upside. Mastering these requirements creates a highly efficient operational pipeline where good governance functions as a massive accelerant for product delivery rather than an administrative handbrake.
Commercial lending and the price of opacity
The mechanics of retail and commercial lending perfectly illustrate the tangible business impact of proper algorithmic oversight.
Consider a scenario where a multinational bank introduces a deep learning framework to process commercial loan applications. This automated system evaluates credit scores, market sector volatility, and historical cash flows to generate an approval decision in a matter of milliseconds. The resulting competitive edge is immediate and obvious, as the institution reduces administrative overhead while clients secure necessary liquidity exactly when they require it.
However, the inherent danger of this velocity resides entirely within the training data. If the deployed model unknowingly utilises proxy variables that discriminate against a specific demographic or geographic area, the ensuing legal consequences are swift and punishing.
Modern regulators demand total explainability and categorically refuse to accept the complexity of neural networks as an excuse for discriminatory outcomes. When an external auditor investigates why a regional logistics enterprise was denied funding, the bank must possess the capability to trace that exact denial directly back to the specific mathematical weights and historical data points that caused the rejection.
Investing capital into ethics and oversight infrastructure is essentially how modern banks purchase speed-to-market. Constructing an ethically-sound and thoroughly vetted pipeline enables an institution to release new digital products without constantly looking over its shoulder out of fear. Guaranteeing fairness from the absolute beginning prevents nightmarish scenarios that involve delayed product rollouts and retrospective compliance audits. This level of operational confidence translates directly into sustained revenue generation while entirely avoiding massive regulatory penalties.
Engineering unbroken information provenance
Achieving this high standard of safety is impossible without adopting a brutal and uncompromising approach toward internal data maturity. Any algorithm merely reflects the information it consumes.
Unfortunately, legacy banking institutions are infamous for maintaining highly fractured information architectures. It remains incredibly common to discover customer details resting on thirty-year-old mainframe systems, transaction histories floating in public cloud environments, and risk profiles gathering dust within entirely separate databases. Attempting to navigate this disjointed landscape makes achieving regulatory compliance physically impossible.
To rectify this, data officers must enforce the widespread adoption of comprehensive metadata management across the entire enterprise. Implementing strict data lineage tracking represents the only viable path forward. For example, if a live production model suddenly exhibits bias against *********-owned businesses, engineering teams require the exact capability to surgically isolate the specific dataset responsible for poisoning the results.
Constructing this underlying infrastructure mandates that every single byte of ingested training data becomes cryptographically signed and tightly version-controlled. Modern enterprise platforms must maintain an unbroken chain of custody for every input, stretching all the way from a customer’s initial interaction to the final algorithmic ruling.
Beyond data storage, integration issues arise when connecting advanced vector databases to these legacy systems. Vector embeddings require massive compute resources to process unstructured financial documents. If these databases are not perfectly synchronised with real-time transactional feeds, the AI risks generating severe hallucinations, presenting outdated or entirely fabricated financial advice as absolute fact.
Furthermore, as we’re currently all too aware, economic environments change at a rapid pace. A model trained on interest rates from three years ago will fail spectacularly in today’s market. Technology teams refer to this specific phenomenon as concept drift.
To combat this, developers must wire continuous monitoring systems directly into their live production algorithms. These specialised tools observe the model’s output in real-time, actively comparing results against baseline expectations. If the system begins to drift outside approved ethical parameters, the monitoring software automatically suspends the automated decision-making process.
Exceptional predictive accuracy means absolutely nothing without real-time observability; without it, a highly-tuned model becomes a corporate liability waiting to explode.
Defending the mathematical perimeter
Of course, implementing governance over financial algorithms introduces an entirely new category of operational headaches for CISOs. Traditional cybersecurity disciplines focus primarily on building protective walls around endpoints and corporate networks. Securing advanced AI, however, requires actively defending the actual mathematical integrity of the deployed models. This represents a complex discipline that most internal security operations centres barely understand.
Adversarial attacks present a very real and present danger to modern financial institutions. In a scenario known as a data poisoning attack, malicious actors subtly manipulate the external data feeds that a bank relies upon to train its internal fraud detection models. By doing so, they essentially teach the algorithm to turn a blind eye to specific and highly-lucrative types of illicit financial transfers.
Consider also the threat of prompt injection, where attackers utilise natural language inputs to trick generative customer service bots into freely handing over sensitive account details. Model inversion represents another nightmare scenario for executives, occurring when outsiders repeatedly query a public-facing algorithm until they successfully reverse-engineer the highly confidential financial data buried deep within its training weights.
To counter these evolving threats, security teams are forced to bury zero-trust architectures deep within the machine learning operations pipeline. Absolute device trust becomes non-negotiable. Only fully-authenticated data scientists, working exclusively on locked-down corporate endpoints, should ever possess the administrative permissions required to tweak model weights or introduce new data to the system.
Before any algorithm touches live financial data, it must successfully survive rigorous adversarial testing. Internal red teams must intentionally attempt to break the algorithm’s ethical guardrails using sophisticated simulation techniques. Surviving these simulated corporate attacks serves as a mandatory prerequisite for any public deployment.
Eradicating the engineering and compliance divide
The highest barrier to creating safe AI is rarely the underlying software itself; rather, it is the entrenched corporate culture.
For decades, a very thick wall separated software engineering departments from legal compliance teams. Developers were heavily incentivised to chase speed and rapid feature delivery. Conversely, compliance officers chased institutional safety and maximum risk mitigation. These groups typically operated from entirely different floors, used different software applications, and followed entirely different performance incentives.
That division has to come down. Data scientists can no longer construct models in an isolated engineering vacuum and then carelessly toss them over the fence to the legal team for a quick blessing. Legal constraints, ethical guidelines, and strict compliance rules must dictate the exact architecture of the algorithm starting on day one. Leaders need to actively force this internal collaboration by establishing cross-functional ethics boards. Banks should pack these specific committees with lead developers, corporate counsel, risk officers, and external ethicists.
When a particular business unit pitches a new automated wealth management application, this ethics board dissects the entire project. They must look past the projected profitability margins to deeply interrogate the societal impact and regulatory viability of the proposed tool.
By retraining software developers to view compliance as a core design requirement rather than annoying red tape, a bank actively builds a lasting culture of responsible innovation.
Managing vendor ecosystems and retaining control
The enterprise technology market recognises the urgency surrounding compliance and is aggressively pumping out algorithmic governance solutions.
The major cloud service providers now bake sophisticated compliance dashboards directly into their AI platforms. These tech giants offer banks automated audit trails, reporting templates designed to satisfy global regulators, and built-in bias-detection algorithms.
Simultaneously, a smaller ecosystem of independent startups offers highly specialised governance services. These agile firms focus entirely on testing model explainability or spotting complex concept drift exactly as it happens.
Purchasing these vendor solutions is highly tempting. Buying off-the-shelf software offers operational convenience and allows the enterprise to deploy governed algorithms without writing heavy auditing infrastructure from scratch. Startups are rapidly building application programming interfaces that plug directly into legacy banking systems, providing instant, third-party validation of internal models.
Despite these advantages, relying entirely on outsourced governance introduces a risk of vendor lock-in. If a bank ties its entire compliance architecture to one hyperscale cloud provider, migrating those specific models later to satisfy a new local data sovereignty law becomes an expensive and multi-year nightmare.
A hard line must be drawn regarding open standards and system interoperability. The specific tools tracking data lineage and auditing model behaviour have to be completely portable across different environments. The bank must retain absolute control over its compliance posture, regardless of whose physical servers actually hold the algorithm.
Vendor contracts require ironclad provisions guaranteeing data portability and safe model extraction. A financial institution must always own its core intellectual property and internal governance frameworks.
By fixing internal data maturity, securing the development pipeline against adversarial threats, and forcing legal and engineering teams to actually speak to one another, leaders can safely deploy modern algorithms. Treating strict compliance as the absolute foundation of engineering guarantees that AI drives secure and sustainable growth.
See also: Ocorian: Family offices turn to AI for financial data insights
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Secure governance accelerates financial AI revenue growth appeared first on AI News.
View the full article
Glia, a customer service platform providing AI-powered interactions for the banking sector, has been named a winner in the Banking and Financial Services Category at the 2026 Artificial Intelligence Excellence Awards. The awards recognises achievements in a range of industries and use cases, spotlighting “companies and leaders moving AI beyond experimentation and into practical, accountable deployment.”
Speaking on the awards, Russ Fordyce, Chief Recognition Officer at Business Intelligence Group commented, “AI has arrived! 2026 is about execution and results. Glia stood out because its work in banking reflects where the market is headed: practical AI that solves real problems, earns trust, and delivers measurable value. The recognition highlights a team that is not participating in the AI shift, but helping define what meaningful progress looks like.”
Glia’s Banking AI platform helps financial institutions navigate security and regulatory risks common in generative AI. It was chosen by a panel of AI experts and analysts as a platform that deploys AI trained precisely for banking workflows. It helps banks and credit unions automate up to 80% of all interactions, according to Glia. For the customer service and member care functions, this can free up time for other tasks, including strengthening client relationships and expanding lending and deposit portfolios; in other words, doing what humans can do and AI can’t.
Dan Michaeli, CEO and co-founder of Glia, said: “The award celebrates the future of banking in an time where AI is everywhere. With consumers in every demographic now using AI to manage their lives, the pressure on financial institutions to provide instant, intelligent service has never been higher.”
“Our platform is designed to help banks and credit unions lead this transition, using secure, banking-specific AI to amplify their efficiency while protecting the human connection that defines their brand,” he said.
Glia has enjoyed positive business momentum recently, with the company announcing recently it will be the first to contractually promise to resist AI hallucinations and circumvent prompt injections for its clients’ use of the platform.
As AI becomes increasingly complex, particularly in financial institutions, Glia’s focus on AI safety provides a model that banks and credit unions might rely on to help them use AI effectively and securely.
(Image source: “Space Invaders does cones and safety barriers” by Gene Hunt is licensed under CC BY 2.0. To view a copy of this license, visit [Hidden Content])
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Glia wins Excellence Award for safer AI in banking appeared first on AI News.
View the full article
When Pew Research Centre analysed 68,879 Google searches in March 2025, one finding stood out: users who encountered an AI-generated summary clicked on a traditional result just 8% of the time. Those who didn’t see a summary clicked nearly twice as often, at 15%. A quarter of users who saw an AI summary ended their session without clicking on anything at all.
That gap tells you something important about where brand discovery is heading. With generative AI platforms like ChatGPT now pulling in 5.72 billion monthly visits (according to SimilarWeb data from January 2026), brands already know AI search matters. The more pressing question is whether your content is structured for the two distinct ways AI retrieves and presents information. SimilarWeb’s framework for AEO vs GEO draws a useful line between these approaches, and it’s one worth understanding before your competitors do.
Where your clicks went and why they’re not coming back
People are searching more than ever. They’re just not clicking.
BrightEdge reported in May 2025 that Google search impressions climbed 49% in the year following the launch of AI Overviews. Over that same *******, click-throughs dropped nearly 30%. Seer Interactive’s September 2025 study, covering 25.1 million organic impressions in 42 organisations, found the decline was even steeper for queries triggering AI Overviews specifically:
Organic CTR fell 61%, from 1.76% to 0.61%
Paid CTR dropped 68%, from 19.7% to 6.34%
Even queries without AI Overviews saw organic CTR decline 41% year-over-year
By March 2025, one in five Google searches produced an AI summary (Pew Research Centre)
Gartner predicted in early 2024 that traditional search volume would fall 25% by 2026. The exact figure remains debatable, but the direction is clear. Impressions are up. Engagement with links is collapsing. The answer itself has become the destination, and the brands inside that answer are the ones getting noticed.
Getting cited by the machine
This is where the AEO vs GEO distinction earns its weight.
Answer Engine Optimisation (AEO) is about structuring content so AI systems can extract a clean, direct answer. Think featured snippets, People Also Ask boxes, voice assistant results. It’s tactical: question-based headings, answer-first paragraphs of 40 to 80 words, FAQ and HowTo schema markup. If someone asks a specific question and your content gives the clearest answer, AEO is what gets you cited at snippet level.
Generative Engine Optimisation (GEO) operates at a broader level. It’s about making your brand a trusted source for RAG-powered platforms (ChatGPT, Perplexity, Gemini) that synthesise answers from multiple sources. GEO involves semantic content clusters, entity-rich data, multimodal assets and building domain authority through co-mentions in third-party sites, directories and publications.
Here’s the part most brands are missing: you can win the featured snippet and still be completely absent from a ChatGPT response. McKinsey’s AI Discovery Survey (August 2025, surveying 1,927 consumers) found that a brand’s own website accounts for only 5 to 10% of the sources AI search platforms reference. The other 90% comes from publishers, user-generated content, affiliate sites and review platforms. So your AEO might be flawless on Google, while your GEO presence in the wider web remains thin.
Worth noting: BrightEdge found that 89% of AI Overview citations come from results ranked beyond position 100. Traditional ranking position is becoming less relevant than content structure and authority signals.
The brands that get cited will be the brands that get chosen
The data on citation advantage is hard to ignore. Seer Interactive’s study found that brands cited in AI Overviews earn 35% more organic clicks and 91% more paid clicks compared to those left out of the summary entirely.
The investment case is building, too. According to Conductor research reported by MarTech in February 2026, 32% of digital marketing leaders now rank GEO as their top priority for the year, and 97% report positive results from their efforts so far. An average of 12% of 2025 digital budgets went to GEO initiatives. Perhaps more telling, 93% of leaders are building these abilities in-house, treating AI search visibility as too strategically important to outsource.
High-maturity organisations are already spending nearly twice as much on GEO as their lower-maturity peers. That gap will be difficult to close once the default answers are set.
If 44% of consumers already prefer AI-powered search as their primary source of insight (McKinsey), and your brand doesn’t appear in those AI-generated responses, where does that leave you in the buying process?
The new front door is already open
AEO and GEO are distinct in their mechanics, but they serve the same purpose: making your brand the one AI systems trust, retrieve and cite. The practical starting point is straightforward. Audit your current AI visibility by prompting the major platforms with questions your customers ask. Identify where you appear, where you don’t and what sources are being cited instead. Then layer AEO (structured answers, schema, question-led content) with GEO (semantic depth, third-party co-mentions, multimodal assets) on top of your existing SEO foundations.
The stakes are rising. As generative AI moves beyond summaries and toward agentic systems that act on users’ behalf (booking, purchasing, recommending), the brands AI cites will increasingly be the brands AI chooses. If your content strategy still measures success by clicks alone, what happens when the click becomes optional?
(Image source: Bazoom)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. This comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post How AEO vs GEO reshapes AI-driven brand discovery in 2026 appeared first on AI News.
View the full article
As artificial intelligence becomes a driving force in financial prediction, the reliability of its forecasting tools faces increasing scrutiny. Many traders question whether claims of high accuracy translate into consistent results under live market conditions. Understanding how these AI systems are evaluated reveals important distinctions between performance in theory and practice.
Few financial domains are as dependent on accurate prediction as forex trading, where slight changes in exchange rates can have consequences for participants. The surge of AI powered price forecasting tools has brought new abilities, but it has also raised questions about what constitutes meaningful accuracy. Readers in this rapidly evolving landscape of predictive technology seek clarity on how well these tools perform and which factors should inform their assessment of forecasts in live environments.
Scrutinising claims of accuracy in predictive tools
Accuracy claims regarding AI forecasting in currency markets are often presented optimistically, particularly when based on controlled demonstrations. These scenarios typically reflect historical data or optimised backtests, which can differ sharply from the volatility and unpredictability seen in live trading environments. The central issue lies in the gap between demonstration results and how models react to real-time market changes. While technical accuracy metrics are frequently referenced, their practical meaning for financial decision-making can remain ambiguous.
When evaluating the accuracy of AI powered price forecasting tools, it is crucial to clarify what “accuracy” represents in this context. For some, accuracy might mean correctly predicting the direction of currency moves, while for others, it could relate to the exact magnitude or timing of price changes. The complexity of forex, with its fast moving variables and interdependencies, underscores why simplistic accuracy scores rarely provide the full picture. Professional users often demand both statistical rigor and domain expertise to interpret results effectively.
Understanding the mechanics behind AI market predictions
AI powered price forecasting tools commonly employ machine learning models specialised for time series prediction. These tools typically use advanced architectures like recurrent neural networks, convolutional neural networks, or transformer-based models designed to capture sequential patterns in financial data. They rely on inputs ranging from historical pricing and trading volumes to macroeconomic indicators and alternative data sources, including geopolitical events or sentiment analysis from news and social media.
There are varied approaches in predictive modeling, with some systems focusing on point predictions that offer specific future prices, while others generate probabilistic forecasts reflecting outcome likelihoods in confidence intervals. The distinction affects how users interpret and trust model outputs. Although probabilistic methods can better accommodate market uncertainty, understanding distributional forecast accuracy and related concepts requires additional expertise. This complexity highlights why headline accuracy figures alone are not sufficient for assessing a system’s practical value.
Evaluating model performance with robust accuracy metrics
Practitioners typically assess AI powered price forecasting tools using a range of evaluation metrics, each shedding light on different facets of prediction quality. Directional accuracy measures whether forecasts correctly predict upward or downward movement of currency pairs, while metrics like mean absolute error or root mean squared error focus on the magnitude of prediction errors. Calibration, which reflects how well predicted probabilities align with actual market occurrences, adds another important dimension.
Meaningful assessment requires benchmarks and rigorous out-of-sample testing, because models effective on past data may not remain reliable as markets change. Overfitting, where models treat noise as signal, can cause high-scoring tools to lose effectiveness once deployed. Similarly, regime shifts and nonstationarity in forex can quickly undermine predictive accuracy, highlighting the importance of ongoing monitoring and validation. It is recognised that participants benefit from understanding both the strengths and limitations of these tools before integrating them into operational processes.
Navigating real world frictions and effective risk controls
When AI powered price forecasting tools are integrated into live strategies, various real world frictions become significant. Issues like latency – the delay between signal and execution – with slippage, spread widening, and inconsistent execution quality, may degrade results observed in backtesting. And, data quality concerns and the risk of look ahead bias present ongoing challenges, particularly if datasets inadvertently include future information unavailable at decision time. As algorithmic signals become more prevalent, financial markets may adapt, reducing the effectiveness of commonly used forecasting techniques.
Effective deployment requires a blend of quantitative insight and robust risk management. Rather than relying solely on single-point forecasts, applying confidence intervals and scenario analysis can yield greater operational stability. Position sizing rules and drawdown controls, with continuous stress testing during volatile periods, help mitigate the effects of erroneous predictions. Ongoing review and adaptation, grounded in an understanding of model limitations and maintained with human oversight, are essential for the sustainable application of AI powered price forecasting tools in currency markets.
(Image source: Bazoom)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Assessing AI powered price forecasting tools in currency markets appeared first on AI News.
View the full article
A developer of API and AI connectivity technologies, Kong, has announced that Bruce Felt has joined it as CFO. Felt is a seasoned finance leader who brings experience guiding enterprise software companies through their growth phases, including several IPOs, acquisitions, and global expansions.
Mr. Felt has led finance organisations from early-stage environments to significant global enterprises. Over his career, he’s taken three companies public as CFO: FullTime Software, SuccessFactors, and Domo. At Domo, a cloud-based analytics and business intelligence software company, he helped scale the business and led the company to its public offering.
Bruce Felt, new CFO at Kong. Source: AZK Media
Augusto Marietti, chief executive officer and co-founder of Kong said:”Bruce has repeatedly helped high-growth software companies scale through transformative periods, pairing operational discipline with strategic insight and several crossings into public markets. As Kong continues to expand its leadership in API and AI connectivity, his experience building durable, globally scaled organisations will be a unique asset in our next journey.”
“He brings the right mix of operational rigor and public company experience, while keeping a growth-oriented profile. We’re extremely excited to welcome Bruce onto the Kong team, and I look forward to partnering with and learning from him.”
Bruce Felt serves on the boards of directors of several organisations, including Veradigm, Human Interest, Betterworks, and Cambium Networks. He has held board and audit committee leadership roles at public and private companies.
(Image source: Pixabay under licence.)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Kong names Bruce Felt as chief financial officer appeared first on AI News.
View the full article
Banking house JPMorgan Chase is asking its roughly 65,000 engineers and technologists to use AI tools as part of their regular workflow. Business Insider reported that managers are tracking how often staff use these tools. That use may also influence performance reviews.
The report states employees are encouraged to use tools like ChatGPT and Claude Code when writing code, reviewing documents, or handling routine tasks. Internal systems then classify workers based on their level of use. Some are labelled “light users,” while others fall into a “heavy user” category.
JPMorgan has been using in fraud detection and risk analysis. What stands out here is not the technology itself, but how it is being woven into day-to-day expectations for staff.
According to internal materials cited by Business Insider, managers are paying close attention to how employees use AI tools.
JPMorgan shows AI adoption in banks
Many companies have spent the past two years rolling out AI tools in departments. In most cases, adoption has been uneven. Some teams experiment heavily, while others stick to existing workflows.
JPMorgan is treating AI as a standard part of the job. That creates a more uniform level of adoption in teams. In the past, performance reviews focused on output and accuracy. Now, they may also include how effectively employees use AI tools to reach those results.
That raises a practical question for large organisations. If AI can reduce the time needed for certain tasks, should employees be expected to produce more work in the same amount of time?
Keeping pace with internal change
By tracking use, the bank may be trying to avoid a familiar problem in enterprise software rollouts. Tools are deployed, but adoption is slow, limiting their impact. Making AI part of performance reviews creates a stronger incentive to engage with the technology. It also suggests that AI literacy is becoming a baseline skill, similar to how spreadsheets or code tools became standard over time.
New challenges include employees feeling pressure to use AI even in cases where it does not clearly improve the outcome. There is also the matter of how to measure “good” use, as opposed to simply frequent use.
JPMorgan’s AI risks and efficiency gains
Banks operate in a regulated environment, where introducing AI into more workflows increases the need for oversight.
Tools like ChatGPT and Claude Code can help summarise information or generate drafts, but they can also produce incorrect or incomplete results. That means employees still need to verify outputs before using them in decision-making or client-facing work.
JPMorgan has developed internal controls for AI systems in areas like trading and risk. Expanding use in a broader group of employees may require similar safeguards, creating a situation for the bank in which it wants to improve efficiency, but also needs to ensure that heavier AI use does not introduce new risks.
Other financial institutions are likely watching closely. If tying AI use to performance leads to measurable gains in productivity, similar models may spread in the sector.
The bank’s approach may reshape how companies hire and train employees, and skills like prompt writing and output checks could become part of standard job requirements. JPMorgan’ approach suggests that this change is already underway, at least in banking.
(Photo by IKECHUKWU JULIUS UGWU)
See also: RPA matters, but AI changes how automation works
Want to experience the full spectrum of enterprise technology innovation? Join TechEx in Amsterdam, California, and London. Covering AI, Big Data, Cyber Security, IoT, Digital Transformation, Intelligent Automation, Edge Computing, and Data Centres, TechEx brings together global leaders to share real-world use cases and in-depth insights. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post JPMorgan begins tracking how employees use AI at work appeared first on AI News.
View the full article
RPA (robotic process automation) is a practical and proven way to reduce manual work in business processes without AI systems. By using software bots to follow fixed rules, companies can automate repetitive tasks like data entry and invoice processing, and to a certain extent, report generation. Adoption grew quickly in many sectors, especially in finance, operations, and customer support.
In recent years the technology has matured. While RPA is still used, business processes can become more complex. Many systems handle unstructured data, like messages and documents. Rule-based automation struggles to handle these inputs, since it depends on predefined steps and structured formats. RPA works best in stable environments where processes do not change often. When conditions change or inputs vary, bots can fail or need updating, adding maintenance overhead and reducing the value of automation over time.
Gartner has pointed to more adaptive automation systems on the market, designed to handle variation and uncertainty, combining automation with machine learning or language models, allowing them to process a broader set of inputs.
From RPA rules to AI-driven automation
AI has changed how companies think about automation, as systems from vendors already known in the RPA space, like Appian and Blue Prism, can now interpret context and adjust their activities, especially relevant for tasks that involve text or images.
Large language models’ ability to summarise documents and extract important details, and respond to queries in natural language offers automation in areas previously difficult to manage. McKinsey & Company research suggests generative AI could automate decision-making and communication work tasks, not routine data handling.
The change does not replace automation, but rather modifies it. Rather than building chains of rules, businesses could use AI to handle variations in input media. Automation becomes more flexible, with systems able to adjust to different inputs without reconfiguration.
That’s the theory. AI systems produce inconsistent outputs, and their behaviour is not predictable. Firms can combine AI with existing automation tools, using each where it fits best. Getting the balance right – intelligent automation – is a hot topic at industry events and on the pages of the RPA and AI media outlets.
Where RPA still fits with AI
Despite these changes, RPA remains relevant in many settings. Tasks that involve structured data and stable workflows still benefit from rule-based automation. Common examples include payroll processing and compliance checks, as well as system integrations.
In these circumstances, RPA’s predictability can be an advantage. Bots follow defined steps and produce consistent results, which is useful in regulated environments. Financial reporting and auditing processes, for example, frequently require strict control and traceability.
Rather than being replaced, RPA is often used with AI. Automation workflows may begin with AI systems that interpret input, then pass structured data to RPA bots for execution. The combination allows companies to extend automation without discarding existing systems.
Blue Prism and the change toward intelligent automation
Vendors that built their business around RPA are adapting to this change. Blue Prism, now part of SS&C Technologies, has expanded its focus to include what it describes as intelligent automation. This approach combines RPA with AI tools capable of processing more complex inputs.
Platforms combine automation with abilities like document processing and decision support, frequently through integrations with AI tools.
The move toward AI-enabled automation also changes how platforms get used. Workflows bring together data sources and decision points, along with execution steps in a single process.
A gradual transition, not a full replacement
Many organisations continue to rely on existing RPA systems, especially where processes are stable and well understood. Replacing these systems would take time and money, which may not always be justified.
Instead, the transformation is gradual. Companies can add AI abilities to extend what automation can handle, while RPA is still in place for tasks where it still works well. This may change how automation is designed and deployed over time, but rule-based systems will remain necessary.
See also: AI agents enter banking roles at Bank of America
Want to experience the full spectrum of enterprise technology innovation? Join TechEx in Amsterdam, California, and London. Covering AI, Big Data, Cyber Security, IoT, Digital Transformation, Intelligent Automation, Edge Computing, and Data Centres, TechEx brings together global leaders to share real-world use cases and in-depth insights. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post RPA matters, but AI changes how automation works appeared first on AI News.
View the full article
To gain financial data insights, the majority of family offices now turn to AI, according to new research from Ocorian. The global study reveals 86 percent of these private wealth groups are utilising AI to improve their daily operations and data analysis.
Representing a combined wealth of $119.37 billion, these organisations want machine learning to modernise their workflows. The technology offers practical benefits for institutions handling complex portfolios, particularly in detecting anomalies, streamlining reporting, and navigating strict regulatory frameworks.
Securing financial data insights via AI and system governance
Implementing these tools requires careful alignment with existing enterprise architectures. Financial institutions frequently rely on major cloud ecosystems, such as Microsoft Azure or Google Cloud, to provide the necessary computing power and security protocols for advanced data processing. By using these platforms, operations teams can deploy machine learning models that identify potential fraud patterns or compliance breaches much faster than manual reviews allow.
While 26 percent of surveyed wealth executives strongly agree that AI will reshape administration and boost performance within the next year, 72 percent expect the broader effects to materialise over a two to five-year horizon.
This cautious timeline reflects the reality of integrating complex algorithms into highly-regulated environments. Integrating new systems without disrupting daily client services presents a major challenge. Legacy data architectures often require heavy re-engineering before they can fully support predictive analytics.
Michael Harman, Commercial Director for the *** and Channel Islands at Ocorian, said: “Family offices are gradually adopting AI and technology as part of their operations and are particularly using it for data insights … there is a realisation that it will have a major impact and family offices need to start exploring the sector and will need support in making the transition.”
Balancing operational upgrades with capital exposure
Despite high operational adoption rates, direct capital allocation into the AI sector remains low. Only seven percent of respondents across 16 territories – including the ***, US, UAE, and Singapore – are currently seeking direct investment opportunities in such technology firms.
This current hesitation highlights a preference for using proven enterprise solutions rather than absorbing the venture-style risks associated with emerging startups. Leaders are focused on immediate operational stability and verifiable returns on investment.
However, this dynamic is likely to change rapidly over the next three years, as 74 percent of these organisations expect to increase their investments in digital assets. Within that group, 20 percent plan to increase their financial commitment to the sector dramatically.
Outsourcing the technical burden to established service providers allows institutions to benefit from enhanced fraud detection and compliance monitoring without directly managing the algorithmic infrastructure. Success will depend on establishing clean data pipelines and ensuring cross-functional teams understand how to interpret algorithmic outputs for risk assessment.
By prioritising secure and scalable cloud platforms, and focusing on specific operational pain points like regulatory reporting, financial leaders can effectively use these AI capabilities to bolster their data insights while maintaining the necessary oversight required in modern wealth management.
See also: AI agents enter banking roles at Bank of America
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Ocorian: Family offices turn to AI for financial data insights appeared first on AI News.
View the full article
AI agents are starting to take on a more direct role in how financial advice is delivered, as large banks move beyond internal tools and into systems that support real client interactions.
Bank of America is now deploying an internal AI-powered advisory platform to a subset of financial advisors, rolled out to around 1,000 financial advisers, according to Banking Dive. The move is one of the clearer early examples of how AI is being used in core banking roles rather than back-office tasks or limited pilots. It also reflects a broader shift across the industry, where AI is moving from basic assistance to systems that can support decision-making in real time.
The platform is based on Salesforce’s Agentforce, which enables the creation of AI agents to handle tasks. It is designed to help advisors handle client queries and prepare recommendations. It can also help manage daily workflows. According to Banking Dive, the system is part of a wider push among major banks to test how AI agents can work alongside human staff rather than operate as standalone tools.
Bank of America has been expanding its use of AI across the business. The bank has said its virtual assistant Erica handles work equivalent to about 11,000 employees, while all 18,000 of its software developers use AI coding tools that have improved productivity by around 20%, according to Banking Dive. These figures give a sense of how widely AI is already embedded across different parts of the organisation.
AI agents move closer to financial decision-making
This approach differs from earlier deployments of AI in banking, which focused mainly on chatbots or internal productivity tools. In those cases, AI was used to answer simple questions or automate routine tasks. The newer systems are built to handle more complex work, including analysing client data and suggesting next steps.
That shift brings AI closer to the core of financial decision-making. Instead of acting as a support layer, the technology is now embedded within the advisory process itself.
Other large banks are moving in a similar direction. The same Banking Dive report notes that firms such as JPMorgan, Wells Fargo, and Goldman Sachs are also testing AI tools aimed at improving productivity and helping staff in client-facing roles, though these efforts vary and are not always focused on advisor-specific AI agent systems. While each bank is taking a different approach, the common goal is to increase output without expanding headcount at the same rate.
Early data suggest these tools can improve efficiency, though results vary. In some cases, banks report gains in how quickly advisors can access information or prepare for meetings, based on industry reporting and early deployment feedback cited by Banking Dive. At the same time, there are ongoing concerns about accuracy and oversight, especially when AI systems are used to suggest financial decisions.
A wider pattern is emerging across financial services. Many institutions are investing in AI, but they are doing so in a controlled way, often limiting deployment to specific teams or use cases. The goal is to test how the technology performs in real settings before expanding further.
Some analysts remain cautious about how quickly AI is changing banking. Wells Fargo analyst Mike Mayo wrote that recent developments have yet to produce major new products, describing the current phase as “a little boring from a product standpoint,” according to Banking Dive.
Human oversight remains central
Bank of America’s rollout stands out because of its scale and placement. Financial advisors sit at the centre of the bank’s relationship with clients, particularly in wealth management. Introducing AI into that role suggests a growing level of trust in the technology. It also shows a willingness to let it influence how advice is formed and delivered.
At the same time, the system is not replacing advisors. Instead, it is meant to work alongside them. Human monitoring remains an essential part of the process, particularly when dealing with complex financial decisions or high-value clients. Industry executives also acknowledge that AI is unlikely to completely replace expert roles, particularly in complex financial workflows where context and judgement still matter.
This hybrid model is becoming more common across the sector. Rather than removing people from the loop, banks are trying to combine human judgement with machine-generated insights. Some firms are starting to treat AI as a part of the workforce rather than a tool, with staff expected to work alongside these systems on day-to-day tasks.
Progress comes with limits and trade-offs
There are also practical challenges. AI systems depend on clean, structured data, which is not always easy to achieve in large organisations with legacy systems. Integration with existing tools can take time, and staff may need training to use new systems effectively.
Regulation adds another layer of complexity. Financial institutions must ensure that AI-driven recommendations meet compliance standards. They must also be able to explain them if questioned by regulators. This requirement may limit the amount of autonomy provided to AI systems, particularly in areas like lending or investment advice.
Despite these constraints, banks are starting to move beyond experimentation and into operational use, even if progress remains uneven. Some estimates imply that up to one-third of banking jobs, or parts of those roles, could eventually be handled by AI, though timelines remain unclear.
The introduction of AI agents into advisory roles also raises questions about how the job itself may change. If systems can handle more of the analytical work, advisors may spend more time on client relationships and less on preparation. Over time, this could shift the skills required for the role.
At the same time, reliance on AI introduces new risks. Errors in data or model output could affect recommendations, and overreliance on automated systems may reduce critical review by human staff. These issues are still being studied as deployments expand.
What sets the current phase is not just the technology, but where it is being used. Moving AI into frontline roles suggests that banks regard it as a tool for shaping outcomes rather than simply improving efficiency behind the scenes.
Bank of America’s rollout offers a view into how that transition may play out. It shows a large institution testing how far AI can be integrated into everyday work, while still keeping human oversight in place.
As more banks follow a similar path, the focus is likely to shift from whether AI should be used to how it should be managed once it becomes part of core operations.
See also: Visa prepares payment systems for AI agent-initiated transactions
Want to learn more about AI and big data from industry leaders? Check outAI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post AI agents enter banking roles at Bank of America appeared first on AI News.
View the full article
Finance leaders are automating their complex workflows by actively adopting powerful new multimodal AI frameworks.
Extracting text from unstructured documents presents a frequent headache for developers. Historically, standard optical character recognition systems failed to accurately digitise complex layouts, frequently converting multi-column files, pictures, and layered datasets into an unreadable mess of plain text.
The varied input processing abilities of large language models allow for reliable document understanding. Platforms such as LlamaParse connect older text recognition methods with vision-based parsing.
Specialised tools aid language models by adding initial data preparation and tailored reading commands, helping structure complex elements such as large tables. Within standard testing environments, this approach demonstrates roughly a 13-15 percent improvement compared to processing raw documents directly.
Brokerage statements represent a tough file reading test. These records contain dense financial jargon, complex nested tables, and dynamic layouts. To clarify fiscal standing for clients, financial institutions require a workflow that reads the document, extracts the tables, and explains the data through a language model, demonstrating AI driving risk mitigation and operational efficiency in finance.
Given these advanced reasoning and varied input needs, Gemini 3.1 Pro is arguably the most effective underlying model currently available. The platform pairs a massive context window with native spatial layout comprehension. Merging varied input analysis with targeted data intake ensures applications receive structured context rather than flattened text.
Building scalable multimodal AI pipelines for finance workflows
Successful implementation requires specific architectural choices to balance accuracy and cost. The workflow operates in four stages: submitting a PDF to the engine, parsing the document to emit an event, running text and table extraction concurrently to minimise latency, and generating a human-readable summary.
Utilising a two-model architecture acts as a deliberate design choice; where Gemini 3.1 Pro manages complex layout comprehension, and Gemini 3 Flash handles the final summarisation.
Because both extraction steps listen for the same event, they run concurrently. This cuts overall pipeline latency and makes the architecture naturally scalable as teams add more extraction tasks. Designing an architecture around event-driven statefulness allows engineers to build systems that are fast and resilient.
Integrating these solutions involves aligning with ecosystems like LlamaCloud and Google’s GenAI SDK to establish connections. However, processing pipelines rely entirely on the data fed into them.
Of course, anyone overseeing AI deployments for workflows as sensitive as finance must maintain governance protocols. Models occasionally generate errors and should not be relied upon for professional advice. Operators must double-check outputs before relying on them in production.
See also: Palantir AI to support *** finance operations
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Automating complex finance workflows with multimodal AI appeared first on AI News.
View the full article
Evidence cited in an eBook titled “AI Quantum Resilience”, published by Utimaco , shows organisations consider security risks as the leading barrier to effective adoption of AI on data they hold.
AI’s value depends on data amassed by an organisation. However, there are security risks to building models and training them on that data. These risks are in addition to better-publicised threats to intellectual property that exist around the point of inference (prompt engineering, for example).
The eBook’s authors state that organisations need to manage threats throughout their AI development and implementation processes. At the same time, companies can and should prepare to change their security protocols, changes that will become mandatory if quantum computing-powered decryption tools become easily available to bad actors.
Utimaco lists three areas under threat:
Training data can be manipulated by bad actors, degrading model outputs in ways are hard to detect,
Models can be extracted or copied, eroding intellectual property rights,
Sensitive data used during training or inference can be exposed.
Current public key cryptography will become vulnerable in the next ten years, the report’s authors attest; a ******* in which capable quantum systems may emerge. Regardless of the timescale, it’s thought that better organised groups currently collect encrypted data and store it to decrypt when or if quantum facilities become available. Any dataset with long-term sensitivity, including model training data, financial records, or intellectual property, may require protection against future decryption, therefore, Utimaco says.
A migration to quantum-resistant cryptography will affect protocols, key management, system interoperability, and performance, so any migration is likely to take several years. The report’s authors suggest what they term ‘crypto-agility’, which it defines as changing cryptographic algorithms without redesigning underlying systems. ‘Crypto-agility’ is based on the principle of hybrid cryptography – combining established algorithms with post-quantum methods, such as those suggested by NIST.
The eBook’s authors concur that cryptography on its own doesn’t address all possible areas of risk. It advocates the use of hardware-based trust devices that can isolate cryptographic keys and sensitive operations from normal working environments.
If companies are developing their own AI tools and processes, protection on that basis should extend throughout the AI lifecycle, from data ingestion through to training, model deployment, and inference in production. Hardware keys used to encrypt data and sign models can be generated and stored inside a boundary. Model integrity can then be verified before deployment, and sensitive data processed during inference remains protected.
Hardware-based enclaves isolate workloads so that even system administrators with sufficient privileges can’t access any of the data being processed. Hardware modules can verify that the data enclave is in a trusted state before releasing keys – a process of external attestation – helping create a ‘chain of trust’ from hardware to application.
Hardware-based key management produces tamper-resistant logs covering access and operations to support compliance frameworks such as the EU AI Act.
Many of the risks inherent in AI systems are well known if not already exploited. The risk from quantum computing’s ability to decrypt data currently considered safe is less immediate, but the implications should affect data and infrastructure decisions made today, Utimaco states. It advocates:
A strengthening of controls throughout the AI development and deployment lifecycle,
The introduction of ‘crypto-agility’ to allow transition to post-quantum security,
Establishing hardware-based trust mechanisms wherever high-value assets are in play.
(Image source: “Scanning electron micrograph of an apoptotic HeLa cell” by National Institutes of Health (NIH) is licensed under CC BY-NC 2.0. To view a copy of this license, visit [Hidden Content])
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Securing AI systems under today’s and tomorrow’s conditions appeared first on AI News.
View the full article
*** authorities believe improving efficiency across national finance operations requires applying AI platforms from vendors like Palantir. The country’s financial regulator, the FCA, has initiated a project leveraging AI to identify illicit activities.
The FCA is currently testing the Foundry platform from Miami-based software vendor Palantir. This three-month pilot costs upwards of £30,000 per week and focuses on mining the regulator’s internal data lake. The objective centres on detecting money laundering, insider trading, and fraud across the 42,000 financial services businesses under the FCA’s supervision.
Navigating unstructured data lakes
Traditional oversight methods struggle with the sheer volume of information generated by modern markets. AI platforms excel at parsing unstructured intelligence, which regulators gather during investigations into harmful activities like human trafficking and the narcotics trade.
The information fed into these systems spans from highly-confidential internal files and reports on problematic companies to consumer ombudsman complaints. Machine learning tools digest audio recordings from phone calls, social media activity, and email archives.
Uncovering patterns within such a vast array of inputs helps direct enforcement resources exactly where they are needed most. Industry experts note a historical under-exploitation of the intelligence housed within regulatory bodies, making advanced analytics a valuable tool for tackling financial crimes.
When validating AI models, there is often a debate about the merits of synthetic information versus live environments. While standard guidelines encourage using artificial datasets for preliminary testing, the ***’s finance regulatory authority determined that evaluating AI software like Palantir’s required actual operational inputs.
Expanding into national security operations
This public sector adoption extends well beyond financial compliance. In September 2025, the *** government established an AI partnership with Palantir aimed at accelerating military decision-making and targeting capabilities. Palantir plans to invest up to £1.5 billion to establish London as its European defence headquarters, an initiative expected to generate up to 350 jobs.
As businesses evaluate these platforms, the defence sector provides a high-stakes testing environment for data fusion. Military planners utilise these tools to consolidate open-source and classified intelligence, rapidly generating options to neutralise enemy targets. This forms an element of the Digital Targeting Web, which relies on a diverse supplier ecosystem.
Palantir and the military will collaborate on identifying opportunities worth up to £750 million over a five-year *******. To foster broader ecosystem growth, the defence agreement includes provisions for mentoring local startups, assisting smaller British technology firms with expanding into US markets on a pro-bono basis.
Deploying private AI like Palantir’s in *** finance operations
CDOs deploying AI solutions often struggle when balancing processing capabilities with privacy mandates. During an enforcement action, regulators frequently compel companies to surrender extensive records.
Such datasets regularly include the personal bank details, telephone numbers, and complete communication logs of individuals tangentially related to a case. Establishing exact boundaries regarding how a software provider interacts with this intelligence is vital. Before selecting Palantir from a two-vendor shortlist, the FCA claims to have run a competitive procurement process and established strict data protection controls.
To mitigate risks associated with information exposure, the FCA structured its agreement with Palantir so the vendor acts strictly as a data processor. Under this arrangement, the software provider operates solely upon instruction. The regulatory agency maintains exclusive possession of encryption keys for the most classified files, and all hosting and storage remain securely within the ***.
Similar data sovereignty principles apply to the defence partnership, ensuring military intelligence remains freely available across the Ministry of Defence while entirely under national control.
The financial contract explicitly forbids the vendor from copying the ingested intelligence to train its own commercial products. Once the pilot concludes, the vendor must destroy the information. Any intellectual property generated during the analysis phase automatically belongs to the regulator.
Setting limitations on data retention and processing rights ensures internal security standards remain intact while achieving efficiency gains from deploying private AI from vendors like Palantir to improve the ***’s finance operations.
See also: Visa prepares payment systems for AI agent-initiated transactions
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Palantir AI to support *** finance operations appeared first on AI News.
View the full article
Payments rely on a simple model: a person decides to buy something, and a bank or card network processes the transaction. That model is starting to change as Visa tests how AI agents can initiate payments. New work in the banking sector suggests that, in some cases, software agents may soon take on that role.
A recent example comes from Visa, which is rolling out its “Agentic Ready” programme in Europe to test how financial systems handle AI-initiated transactions. The effort involves collaboration with banks, including Commerzbank and DZ Bank. The aim is to prepare existing payment infrastructure for a scenario where software agents can search for products and make decisions, then complete purchases on behalf of users.
According to information published by Visa and reported by The Paypers, the programme focuses on enabling secure transactions where AI systems act as the initiating party. Instead of a customer confirming a purchase, an AI agent could carry out the task after being given a goal or set of rules.
How transactions begin
Payment systems are built around human identity and intent. A card transaction today depends on verifying that a person has authorised a purchase. If AI agents begin to initiate transactions, banks will need new ways to confirm identity and intent at the system level. That includes deciding how an agent proves it is acting on behalf of a user, and how much autonomy it should have.
In Visa’s model, software agents could handle routine or repeat purchases with limited human input, based on user-defined rules. A system could, for example, monitor supply levels and compare prices, then complete a transaction when certain conditions are met. Reporting from Die Welt and Investing.com says the company sees this as similar in scale to the early change toward online payments, when banks had to adapt to a new type of transaction flow.
Control and compliance
Banks involved in early trials are testing how these ideas work in practice. Commerzbank and DZ Bank are exploring how AI agents can be integrated into existing systems without breaking compliance rules. This includes checks related to fraud, audit trails, and customer consent. These areas are tightly regulated, which means any change to how transactions are initiated must still meet oversight standards.
A RepRisk report found that banks are already dealing with more frequent and costly issues linked to AI. The report states that these incidents can lead to multi-million-dollar losses.
Visa’s work is focused on infrastructure not consumer-facing tools. It’s working on how payment networks should behave when the “customer” is a piece of software. That includes defining how agents are authenticated and how transactions are approved. It also covers how disputes are handled if something goes wrong.
AI and enterprise purchasing
In large organisations, procurement often involves multiple approval steps. AI agents could compress that process by handling routine purchases in set limits. This could reduce manual work, but it also means companies need clear rules about what agents are allowed to do. Without that, the risk of errors or misuse increases.
Large institutions are investing in AI to automate back-office work and reduce costs. Some are also reorganising teams to focus more on data and AI strategy. Regulators are paying closer attention to how AI is used in decision-making, especially in areas like credit and fraud detection.
Taken together, these developments suggest that payments could become one of the first areas where AI agents could act with greater autonomy. Banks will still need to set rules, monitor activity, and handle exceptions. But the day-to-day act of initiating a transaction may, in some cases, require less direct human input.
Visa’s current phase is focused on testing and system design. As AI systems take on more responsibility, financial infrastructure will need to adapt to a new type of user, one that does not hold a card but can still make a purchase.
(Photo by CardMapr.nl)
See also: Goldman Sachs sees AI investment change to data centres
Want to learn more about AI and big data from industry leaders? Check outAI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Visa prepares payment systems for AI agent-initiated transactions appeared first on AI News.
View the full article
The NVIDIA Agent Toolkit is Jensen Huang’s answer to the question enterprises keep asking: how do we put AI agents to work without losing control of our data, our systems, and our liability?
Announced at GTC 2026 in San Jose on March 16, the NVIDIA Agent Toolkit is an open source software stack designed to help enterprises and developers build autonomous AI agents–ones that can perceive, reason, and act on their own, across internal systems, without needing a human to babysit every step.
The timing makes sense. The agent conversation has moved well past the pilot phase. What’s stalling broader deployment isn’t capability–it’s trust. Agents that can take action inside enterprise systems need guardrails, and until now, those have been hard to standardise at scale.
OpenShell and the safety problem
The centrepiece of the toolkit is NVIDIA OpenShell, an open source runtime that enforces policy-based security, network, and privacy guardrails for autonomous agents. In NVIDIA’s terminology, individual agents are called “claws”, and OpenShell is what keeps them in check.
Huang framed the stakes plainly at GTC: “Claude Code and OpenClaw have sparked the agent inflexion point–extending AI beyond generation and reasoning into action. Employees will be supercharged by teams of frontier, specialised, and custom-built agents they deploy and manage.”
That last part is the pitch. The ambition isn’t a single AI assistant; it’s a workforce of specialised agents, each handling a domain, coordinated at scale. OpenShell is the layer that’s supposed to make that deployable without IT teams having heart attacks.
NVIDIA is working with Cisco, CrowdStrike, Google, Microsoft Security, and TrendAI to build OpenShell compatibility into their respective security tools, which signals that this isn’t being positioned as a standalone product, but as infrastructure others build on top of.
The research and cost angle
Also inside the toolkit is NVIDIA AI-Q, an agentic search blueprint built with LangChain. It uses a hybrid architecture–frontier models handle orchestration while NVIDIA’s open Nemotron models do the research-heavy lifting. According to NVIDIA, this approach can cut query costs by more than 50% while still producing accuracy that tops the DeepResearch Bench and DeepResearch Bench II leaderboards.
That cost figure will matter to enterprise buyers who’ve been burned by consumption-based AI pricing that looked manageable in pilots and became a budget problem at scale.
Who’s already on board?
The partner list at GTC was extensive. Adobe, Atlassian, SAP, Salesforce, ServiceNow, Siemens, Cisco, CrowdStrike, Red Hat, Box, Cadence, Cohesity, Dassault Systèmes, IQVIA, and Synopsys are all advancing enterprise AI agents using the NVIDIA Agent Toolkit.
A few specifics stand out. Salesforce is building a reference architecture where employees use Slack as the orchestration layer for Agentforce agents–pulling from data in both on-premises and cloud environments–powered by NVIDIA infrastructure. Atlassian is integrating Agent Toolkit into its Rovo AI strategy across Jira and Confluence. ServiceNow’s “Autonomous Workforce of AI Specialists” is built on the toolkit alongside NVIDIA AI-Q.
And Siemens launched the Fuse EDA AI Agent, which uses NVIDIA Nemotron to autonomously orchestrate workflows across its electronic design automation portfolio, from design conception through manufacturing sign-off.
IQVIA’s deployment numbers offer a real-world data point: the company has already deployed more than 150 agents across internal teams and client environments, including 19 of the top 20 pharma companies.
The ******* shift
What NVIDIA is really doing here is positioning itself not just as the hardware backbone of AI, but as the software infrastructure layer for enterprise agentic deployment. The Agent Toolkit, OpenShell, Nemotron models, AI-Q-these are components of a stack that NVIDIA wants sitting underneath an enormous swath of enterprise software.
Whether that bet pays off depends on how quickly enterprises move from agent experimentation to agent operations. The toolkit is available now on build.nvidia.com, with support across AWS, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure.
See also: AI Expo 2026 Day 1: Governance and data readiness enable the agentic enterprise
Want to learn more about AI and big data from industry leaders? Check outAI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post NVIDIA wants to make enterprise AI agents safe enough to actually deploy appeared first on AI News.
View the full article
Mastercard has developed a large tabular model (an LTM as opposed to an LLM) that’s trained on transaction data rather than text or images to help it address security and authenticity issues in digital payments.
The company has trained a foundation model on billions of card transactions, with the intention of expanding to hundreds of billions in time. The datasets include payment events and associated data such as merchant location, authorisation flows, fraud incidents, chargebacks, and loyalty activity. Mastercard says personal identifiers are removed before the training began, and that the model parses behavioural patterns rather than concern itself with individual identities.
By excluding personal data, the technology reduces privacy risks that may affect other forms of AI in financial services sector. The scale and richness of the data allow the model to infer patterns that are commercially valuable – the company said in a recent blog post – despite the lack of per-user information. Although anonymisation removes signals that could be argued as being useful in the area of risk assessment, Mastercard asserts that using sufficiently large volumes of behavioural data compensates for any loss of rich data.
What is an LTM (large tabular model)?
LTM architecture differs from that of large language models, which are trained on unstructured inputs and work by predicting the next token (typically but inaccurately described as a word) in a sequence. Mastercard’s LTM examines relationships between fields in multi-dimensional data tables, making a definition of the technology closer to that of pure machine learning rather than artificial intelligence.
The large tabular model learns from raw inputs exactly which relationships are predictable, so it can identify anomalous patterns not captured by predefined rules.
The company describes the LTM as an ‘insights engine’ that can be used in existing products, augmenting existing workflows. The operational risk of a model that interacts with customers (often an LLM) differs from that of one that’s part of internal decision-making.
Technical infrastructure for the LTM comes from Nvidia and Databricks, with the former providing the computing platform and Databricks handling data engineering and model development.
Where will we see an LTM in operation?
Cybersecurity at Mastercard is the first area to see active deployment of the tech. Like many institutions, Mastercard operates several fraud detection systems examining transaction data. These require human input at their outset – and ongoing attenuation – to define what constitutes as suspicious behaviour. These might include sudden increases in transaction frequency, or users making purchases in different parts of the world in a small space of time.
Early results indicate improved performance on conventional techniques in specific cases, the company says. It cites the example of high-value, low-frequency purchases which can be flagged as anomalies using traditional models, but the new model appears to be able to distinguish legitimate events more accurately than its counterparts.
The company plans to deploy hybrid systems that combine established procedures with the new model, a degree of caution that reflects the regulatory levels it operates under. It acknowledges that no single model is likely to perform well in all scenarios, so the LTM will take its place among the tools in this sphere.
It’s claimed the model can scan activity on loyalty programmes, be used in portfolio management, and for internal analytics, areas where there are large volumes of structured data. In current operations, companies often deploy many models adapted to each task, but this can involve multiples of training costs and validation and monitoring efforts. A single foundation model that can be fine-tuned for different tasks may simplify processes and keep costs down.
Risk and future plans
There’s a risk to the multi-function LTM approach, of course: A failure in a widely-deployed model could have system-wide consequences, which goes some way to explain Mastercard’s strategy of applying its technology alongside existing detection systems – at least, for the present.
Mastercard hopes to increase the scale of the data used on the model and its overall sophistication. It’s also planning on API access and SDKs to let internal teams build new applications.
The blog post emphasises the data responsibilities the LTM holds, mentioning privacy and transparency, model explainability, and auditability. Regulatory scrutiny of any system that influences credit decisions or fraud outcomes is to be expected in addition to any data practices involved in the LTM’s operation.
Highly structured data, as opposed to text or images, lies at the core of the LTM. Large tabular models may be the start of a new generation of AI systems in core banking and payments infrastructure. Evidence to date remains limited to vendor reports, so any performance claims should not necessarily be regarded as conclusive.
Robustness under adversarial conditions, long-term post-training costs, and regulatory acceptance are all issues on which tabular models may founder or thrive. These factors will determine the pace and extent of adoption, but it’s the area of the table where Mastercard is placing some of its bets at present.
(Image source: “Oversight” by United States Marine Corps Official Page is licensed under CC BY-NC 2.0.)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Mastercard keeps tabs on fraud with new foundation model appeared first on AI News.
View the full article
A report from Autorek, a provider of AI solutions to the insurance industry has produced a report that describes operational drag in companies’ internal processes that not only affect overall efficiency but cause an impediment to the effective implementation of AI in insurance concerns. Insurance Operations & Financial Transformation 2026 draws from a survey of 250 managers in the sector from the *** and US. The survey’s responses paint a picture of connected bottlenecks that include slow settlement processes and data fragmentation. The report also covers the current state of AI deployment in the industry.
Companies surveyed in the sector report persistent structural inefficiencies:
14% of operational budgets are spent correcting manual errors,
22% of those questioned said reconciliation complexity is a significant cause of cost increases,
Around 22% of respondents link inefficiencies to governance and audit risks,
Nearly half of firms operate settlement cycles in excess of 60 days.
Transaction volumes are projected to rise by roughly 29% in the next two years means, the report claims, and OPEX burdens are likely to rise commensurately. The report attributes this to the combination of manual processing, disparate data systems, and the transactional complexity that’s the nature of modern insurance operations. The persistence of such processes, the authors state, is despite its previous publications’ findings being in the public domain for some time.
There is a gap between respondents’ expectations of what AI might deliver and implementation of the technology on the ground. The headline figure is that 82% of firms in the sector expect AI to dominate the industry, yet only 14% of companies have fully-integrated AI in their operations. Six percent of companies report no use of AI at all.
What are the barriers to AI in the insurance sector?
The report identifies legacy system integration, fragmented data, and limited internal expertise as the main issues companies need to address to implement AI. The issue of fragmented data affects data governance frameworks, making the latter similarly piecemeal. The report’s authors cite complex data estates in many companies as the main reason that AI deployments are constrained in the sector.
Firms surveyed managed an average of 17 data sources, and a majority cite this as an issue, one that’s compounded after mergers and acquisitions.
The report’s authors imply AI will affect costs and scalability positively and could address some of the issues firms experience around manual error correction and mistakes in reconciliation processes. The report suggests decision-makers could target reconciliation processes for an initial proving ground for AI, given it’s a boundary-ed, rules-based domain where automation can yield fast positive results.
Any form of automation, AI or deterministic, placed on a fragmented architecture and a fractured data layer may not scale well without a rise in costs. The report highlights the potential for AI in structuring fragmented data sources, and suggests cloud-based, as opposed to in-house AI platforms may be an answer in that respect.
Structural issues
The dichotomy between reconciliation processes (essentially structured workflows) and disparate data sources that need manual nurturing creates complexity that’s measurable in cost and cycle times. This is a situation that persists despite a broad awareness of the issues among those surveyed.
The report asserts that such firms successful in addressing the issues at a structural level will widen the performance gap. Data standardisation and governance precede scalable automation, and eventually, automation will reduce reconciliation costs. AI could address the complexity of fragmented data and software layers that rules-based automation such as RPA (robotic process automation) may not be able to address economically.
The rate at which firms can resolve the data fragmentation issue is dictated by legacy technology and the overheads of day-to-day operations. The extent to which AI deployment could translate into performance gains beyond cost reduction is unclear, but if cost reduction is positive outcome enough, then addressing the structural issues affecting the insurance sector would form a solid basis for AI-powered automation.
(Image source: “Scattered pieces” by Cle0patra is licensed under CC BY-NC-SA 2.0.)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post For effective AI, insurance needs to get its data house in order appeared first on AI News.
View the full article
Trustpilot is reported to be pursuing partnerships with large eCommerce companies as AI-driven shopping gains traction.
In an interview with Bloomberg News [paywall], chief executive Adrian Blair said that AI agents acting on behalf of consumers require lots of information about the businesses they’re willing to interact with. He said the most effective systems will rely on datasets like those held by Trustpilot, adding that the company aims to work with major eCommerce sites to make greater use of its data.
Trustpilot expects its operating margin to reach 30% by 2030, with the improvement linked partly to the use of its content by LLMs. According to Bloomberg, traffic patterns are beginning to reflect this. Click-throughs from AI-based search increased by 1,490% over the past year, thanks in no small part to search giant Google’s decision to make an AI search the default.
Data from Promptwatch indicates that Trustpilot ranked as the fifth most cited domain globally in ChatGPT in January this year.
Blair said that large language models have created a new channel through which Trustpilot content is presented, noting a rise in exposure and referral traffic from LLM-based algorithms.
In February 2026, Amazon and OpenAI announced an agreement to deploy genAI systems on AWS using customised models intended for Amazon’s consumer-facing applications. The arrangement is said to cover infrastructure provision and model development.
Elsewhere, Walmart’s partnership with Google lets users purchase goods inside the Gemini chatbot. Google has similar arrangements with Shopify and other retailers.
Shopify’s Universal Commerce Protocol lets AI agents access product data and take transactions to checkout, so ensuring potential buyers remain on the AI platform (in this case Gemini) rather than navigate to the retailer’s site. Microsoft’s Copilot Checkout collaboration with PayPal falls into the same pattern.
Shopify has pursued similar partnerships including with Microsoft so merchants can sell from a chatbot interfaces. Its recent product updates describe “agentic storefronts” in which transactions take place inside AI interactions. For marketing professionals, the loss of valuable data when shoppers purchase through a third-party proxy is, to varying degrees, balanced by the income from trade via AI platforms.
Amazon currently challenges third-party AI agents accessing its platform without authorisation, and is developing its own assistant to retain control over user data and advertising revenue, according to the Wall Street Journal.
Trustpilot’s Adrian Blair argued in the Bloomberg News interview that user-generated reviews retain value regardless of the involvement of AI in the purchasing process. He said consumers will continue to “have experiences” with businesses, describing Trustpilot’s data set of reviews as a long-term asset whose relevance is increasing.
The company’s shares were affected by a broader decline in software stocks last month, sparked by the media imagining the death of SaaS platforms on the back of claims made by Anthropic.
PYMNTS Intelligence’s report , “How AI Becomes the Place Consumers Start Everything,” describes consumers beginning their product research and shopping on AI platforms, refining their prompts iteratively rather than successive ‘traditional’ searches.
(Image source: “E-Commerce Visa (Test tamron 17-50 2.8)” by Fosforix is licensed under CC BY-ND 2.0.)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Trustpilot partners with AI companies as traditional search declines appeared first on AI News.
View the full article
Artificial intelligence investment is entering a more selective phase as companies and investors look beyond early excitement and focus on the data centre infrastructure required to run AI systems.
Recent analysis from Goldman Sachs suggests the market is moving toward what the firm describes as a “flight to quality.” In practice, investors are paying closer attention to companies that own and operate large data centres and computing infrastructure. Firms offering narrow AI tools or experimental software are receiving less attention.
Goldman Sachs expects spending on AI infrastructure to grow rapidly as companies expand computing capacity for model training and deployment. Hyperscale cloud firms are investing tens of billions of dollars each year in new data centres and computing hardware. Networking systems are also expanding to support this growth.
AI demand is reshaping the data centre market
Goldman Sachs Research estimates that AI workloads could account for about 30% of total data centre capacity in the next two years, as demand for computing power grows in cloud services and enterprise applications. The change reflects how AI tasks differ from traditional cloud workloads. Training large models requires thousands of chips running in parallel for extended periods. Inference, the process of generating responses or predictions, also requires steady computing power when services run.
Cloud providers and AI developers are now expanding data centre capacity at a pace not seen during earlier phases of cloud computing. Infrastructure demand extends beyond computing hardware. Energy supply is becoming a central issue in the AI race.
Goldman Sachs Research estimates that global data centre power demand could rise about 175% by 2030 compared with 2023 levels, driven largely by AI workloads. The firm says this increase would be roughly equal to adding the electricity demand of another top-10 power-consuming country to the global grid. Rising power demand is also pushing utilities and governments to consider new investment in energy infrastructure.
Infrastructure limits are shaping AI strategy
The growing need for power and cooling is influencing where new AI data centres are built. Space requirements are also shaping site selection. Large facilities are often located near stable energy sources and high-capacity fibre networks. Some companies are building AI training clusters in remote areas where land and electricity are easier to secure. The location of data centres can also affect environmental impact. Academic research on AI infrastructure shows that cooling systems and geographic location can influence energy use and water consumption as much as hardware efficiency.
The limits are starting to affect how technology firms plan their AI strategies. Building new models or software is only part of the challenge. Companies must also ensure they have the infrastructure needed to run those systems reliably. In many cases, building that infrastructure takes years.
Construction of large data centres involves complex supply chains. Projects often require land acquisition and grid connections. Many also depend on long-term energy agreements. Shortages of electrical equipment and delays in grid expansion can slow new projects. The constraints help explain why investors are paying more attention to companies that already control large data centre networks.
A selective phase of the AI market
During the first wave of generative AI adoption, many companies saw their market value rise simply by associating themselves with AI. That phase is now beginning to change as investors reassess where AI growth will occur.
Investors are examining which companies have the infrastructure and revenue models needed to support long-term deployment. Data centre operators and chip manufacturers sit near the base of that ecosystem. Their services are required regardless of which AI applications gain traction.
During previous waves of computing growth, companies that built the underlying infrastructure often captured stable revenue. Software platforms, in contrast, rose and fell more quickly. A similar dynamic may now be forming in the AI sector.
Infrastructure expansion also raises new questions. Energy demand and grid capacity are becoming central issues for governments and industry planners. Environmental impact is also drawing closer scrutiny.
In the coming years, the AI economy may depend as much on power plants and cooling systems as it does on algorithms and software. That reality is shaping the next stage of the AI race.
(Photo by Lightsaber Collection)
See also: Goldman Sachs and Deutsche Bank test agentic AI for trade surveillance
Want to learn more about AI and big data from industry leaders? Check outAI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Goldman Sachs sees AI investment shift to data centres appeared first on AI News.
View the full article
The US Treasury has published several documents designed for the US financial services sector that suggest a structured approach to managing AI risks in operations and policy (see subheading ‘Resources and Downloads’ towards the bottom of the link). The CRI Financial Services AI Risk Management Framework (FS AI RMF) comes with a Guidebook [.docx] which gives details of the framework, developed by a collaboration among more than 100 financial institutions and industry organisations, with input from regulators and technical bodies.
The objective of the FS AI RMF is to help financial institutions identify, evaluate, manage, and govern the risks associated with AI systems and let firms continue adopting AI technologies responsibly.
Sector-specific framework
AI systems introduce risks that existing technology governance frameworks don’t address. Risks include algorithmic bias, limited transparency in decision processes, cyber vulnerabilities, and complex dependencies between systems and data. LLMs create concerns because their behaviour can be difficult to interpret or predict. Unlike traditional software, which is deterministic, an AI’s output varies depending on context.
Financial institutions already operate under extensive regulation and there is a raft of general guidance such as the NIST AI Risk Management Framework. However, applying general frameworks to the operations of financial institutions lacks the detail that reflects sector practices and regulatory expectations. The FS AI RMF is being positioned as an extension to the NIST framework, with additional sector-specific controls and practical implementation guidelines in its pages.
The Guidebook explains how firms can assess their current AI maturity and implement controls to limit their risk. Its aim is to promote consistent and responsible AI practices and support innovation in the sector.
Core structure
The FS AI RMF connects AI governance with broader governance, risk, and compliance processes already affecting financial institutions.
The framework contains four main components. The first is an AI adoption stage questionnaire that lets organisations determine the maturity of their AI use. The second is a risk and control matrix, which contains a set of risk statements and control objectives in alignment with adoption stages. The Guidebook explains how to apply the framework, while a separate control objective reference guide provides examples of controls and supporting evidence.
The framework defines a total of 230 control objectives organised according to four functions adapted from the broader NIST AI Risk Management Framework: govern, map, measure, and manage. Each function contains categories and subcategories that describe elements of effective AI risk management and governance.
Assessing AI maturity
The adoption stage questionnaire determines the extent to which an organisation is using AI. Some firms rely on traditional predictive models in limited applications for example, while others deploy AI in core business processes; others just use AI in customer-facing roles.
The questionnaire helps organisations determine where they sit in the spectrum of AI use currently, evaluating factors like the business impact of AI, governance arrangements, deployment models, use of third-party AI providers, organisational objectives, and data sensitivity.
Based on this assessment, organisations are classified into four stages of AI adoption:
initial stage: organisations that have little or no operational AI deployment. AI may be under consideration but is not embedded,
minimal stage: limited AI use in low-risk areas or isolated systems.
evolving stage: organisations running more complex AI systems, including applications that involve sensitive data or external services.
embedded stage: where AI plays a significant role in business operations and decision-making.
These stages help institutions focus their efforts on controls appropriate to their maturity level. A firm at an early stage does not need to implement every control immediately, but as AI becomes more integrated, the framework introduces additional controls to address growing levels of risk.
Risk and control
The control objectives for each AI adoption stage address governance and operational topics including data quality management, fairness and bias monitoring, cybersecurity controls, transparency of AI decision processes, and operational resilience.
The Guidebook provides examples of possible controls and types of evidence institutions can use to demonstrate they’re compliant. Each firm must determine the controls that fit best.
The framework recommends maintaining incident response procedures specific to AI systems and creating a central repository for tracking AI incidents, processes that will help organisations detect failures and improve governance over time.
Trustworthy AI
The framework incorporates principles for trustworthy AI defined as validity and reliability, safety, security and resilience, accountability, transparency, explainability, privacy protection, and fairness. These provide a foundation for evaluating AI systems along their full lifecycle. In simple terms, financial institutions have to ensure AI outputs are reliable, that systems are protected against cyber threats, and that decisions can be explained when they affect customers or have regulatory relevance.
Strategic implications
For senior leaders in financial institutions of any nation, the FS AI RMF offers a guide to integrating AI into existing risk management frameworks. It states the need for coordination in different business functions in the organisation. Technology teams, risk officers, compliance specialists, and business units all need to participate in the AI governance process.
Adopting AI without strengthening governance structures may expose institutions to operational failures, regulatory scrutiny, or reputational damage. Conversely, firms that build clear governance processes will be more confident in deploying AI systems.
The Guidebook frames AI risk management as an evolving entity. As AI technologies develop and regulatory expectations change, institutions will need to update their governance practices and risk assessments accordingly.
For financial sector decision-makers, the message is that AI adoption must progress in step with risk governance. A structured framework such as the FS AI RMF provides a common language and method to manage the evolution.
(Image source: “Law Books” by seychelles88 is licensed under CC BY-NC-SA 2.0.)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post US Treasury publishes AI risk Guidebook for financial institutions appeared first on AI News.
View the full article
NTT DATA has announced an initiative to deliver NVIDIA-powered platforms designed to give organisations a repeatable, production-ready model for scaling AI.
The offering integrates NVIDIA’s GPU-accelerated computing and high-performance networking with NVIDIA AI Enterprise software, including NeMo and NIM Microservices, into a full-stack agentic AI platform that can be deployed in cloud and edge environments. The architecture covers the full AI lifecycle of model training and enterprise application development inside a governed framework.
Abhijit Dubey, CEO of NTT DATA said there is a change in how enterprises approach AI deployment. “By integrating NVIDIA technologies into our enterprise AI factories, we’re giving clients a powerful and secure environment to adopt agentic AI with measurable returns from the start.”
NTT DATA says the enterprise AI factory model addresses a gap that has stalled many AI programmes: the distance between a successful pilot and a production system that runs. The platform is designed to standardise output and reduce the time and cost of moving from proof-of-concept to operational deployment.
Real-world deployments
Three early-adopter cases give a clearer picture of enterprise AI factories. A leading *******-research hospital is using NVIDIA HGX platforms, with NTT DATA and Dell, for advanced radiology analysis and rapid model evaluation to support clinical research workflows.
In automotive manufacturing, a global supplier has reduced production setup time by validating workloads on bare metal before scaling through an AI factory architecture on NVIDIA infrastructure. A third deployment, in technology manufacturing, involves a US-based company using NVIDIA-accelerated simulation and 3D visualisation to validate a next-generation battery production line before physical deployment.
NTT DATA is positioning enterprise AI factories as a domain-specific delivery model, with the NVIDIA stack serving as the common infrastructure underneath sector-by-sector customisation.
NeMo and NIM in an AI factory stack
The technical integration comprises of two NVIDIA components. NVIDIA NeMo is a suite for building agentic AI systems on GPU-accelerated infrastructure. NVIDIA NIM Microservices provide pre-built, GPU-optimised containers with APIs for deploying AI applications. Together, they form what NTT DATA describes as a full-stack, production-ready AI agent platform.
NTT DATA also offers pre-qualified GenAI prototypes built on this stack, which it says reduces complexity and accelerates time to value for clients building sector-specific applications.
John Fanelli, Vice President of Enterprise Software at NVIDIA, said: “Enterprises are now seeking robust, scalable platforms that can successfully transition their AI initiatives from pilot projects to full-scale production.” He said NTT DATA’s AI factory offerings provide clients with domain-specific solutions needed to achieve production-grade enterprise AI.
NTT DATA describes itself as the only global IT services provider active in all three of NVIDIA’s partner tracks: Solution Provider, Cloud Partner, and Global System Integrator Partner Network.
The recent announcement comes as enterprises face rising pressure to show financial returns on AI spending. Governance and domain-specific performance are now the criteria by which enterprise AI investments are judged, and the AI factory model is an attempt to make all three more systematic.
See also: Physical AI is having its moment – and everyone wants a piece of it
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post NTT DATA and NVIDIA bring enterprise AI factories to production scale appeared first on AI News.
View the full article
When OpenAI launched Frontier in February, the announcement was framed as a platform for enterprise AI agents. What it actually signalled was a direct challenge to the revenue architecture that has underpinned the software industry for the better part of two decades.
Frontier is designed to act as a semantic layer across an organisation’s existing systems, connecting data warehouses, CRM platforms, ticketing tools, and internal applications so that AI agents can operate with the same business context a human employee would have. OpenAI describes these agents as “AI coworkers” that can be onboarded, assigned identities, granted permissions, and reviewed for performance.
Early customers include Uber, State Farm, Intuit, and Thermo Fisher Scientific. The commercial ambition behind the platform is not subtle. OpenAI CFO Sarah Friar has stated that enterprise customers currently account for roughly 40% of the company’s revenue, and she aims to increase this figure to closer to 50% by year-end. Frontier is the vehicle.
What Frontier actually does to enterprise workflows
The case for Frontier rests on a problem that CIOs have described consistently through 2025 and into this year: agents deployed in isolation add complexity rather than remove it. Each new agent becomes a point of integration, requiring its own data connections and governance controls, and the result is fragmentation at scale.
OpenAI’s answer is a shared business context. Rather than each agent building its own understanding of how an organisation works, Frontier provides a centralised layer that all agents can reference. Fidji Simo, OpenAI’s CEO of Applications, put it plainly during the launch briefing, drawing on her time running Instacart.
“We spent months integrating each of the ones that we selected. We didn’t even get what we actually wanted, because each tool was good for one use case, but they weren’t integrated or talking to one another, so we were just reinforcing silos upon silos.”
The results OpenAI cites from early deployments are notable. A global investment firm using Frontier agents across its sales process freed up more than 90% of salesperson time previously spent on administrative tasks. A technology customer reported saving 1,500 hours a month in product development. At a major manufacturer, agents compressed a production optimisation process from six weeks to a single day.
Frontier is also deliberately open. It manages agents built by OpenAI, agents built in-house by enterprise teams, and agents from third-party providers, including Google, Microsoft, and Anthropic. That openness is both a design principle and a positioning move: it makes Frontier harder to dismiss as a vendor lock-in play, while expanding the surface area it can govern.
The seat-licence problem nobody wants to say out loud
The deeper concern for incumbents is structural. The per-seat licence model that has made SaaS enormously profitable assumes that software usage maps to headcount. If an AI agent handles the workflow that previously required a human employee logging into Salesforce, the justification for that seat licence weakens. Fortune described it directly: the fear in the market is that platforms like Frontier will make SaaS software “invisible” and consequently less valuable.
Salesforce’s stock has declined more than 27% so far this year, a fall analysts have attributed more to agentic AI disruption fears than to any weakness in its underlying financials. The company’s Q4 FY2026 results were solid. Revenue reached $11.2 billion in the quarter, Agentforce’s annual recurring revenue hit $800 million, and the company closed 29,000 Agentforce deals.
The stock still fell after hours, on guidance that came in below Wall Street’s expectations.
The incumbents are not standing still. Salesforce has introduced what it calls the Agentic Enterprise License Agreement, a fixed-price, all-you-can-eat model for Agentforce that attempts to make consumption more predictable for enterprise buyers.
ServiceNow has moved to consumption-based pricing for some of its AI agent offerings, and in January signed a multiyear agreement with OpenAI to embed frontier model capabilities directly into its platform. Microsoft has introduced consumption-based pricing alongside its per-user model for Copilot Studio.
The pricing pivot is significant. It signals that these companies understand the seat-licence model cannot survive agentic AI unchanged. The question is whether repricing is enough or whether the architecture itself needs to change.
Two bets on where the intelligence layer should sit
The strategic divide in enterprise AI right now runs along a single fault line: should AI agents live inside systems of record, or above them? Salesforce and ServiceNow are betting on the embedded model. They argue that agents are most effective when they sit closest to the data, and that CIOs will trust governance and compliance controls more readily from vendors already managing their workflows.
Marc Benioff, CEO of Salesforce, has described Agentforce as the “operating system for the agentic enterprise.” ServiceNow positions its AI Control Tower as a centralised governance layer for all agents, regardless of where they originate.
OpenAI, and to a similar degree, Anthropic with Claude Cowork, is betting on the overlay model. Frontier sits above existing systems, using open standards to connect them rather than replacing them. The pitch is that enterprises should not have to replatform to get production-grade agents running across their operations.
Both arguments have merit, and enterprises evaluating these platforms will find genuine trade-offs. The embedded approach offers tighter data control and faster time to value within a known ecosystem. The overlay approach offers flexibility and avoids the problem of agents that can only see one vendor’s data.
What the incumbents have that OpenAI does not is decades of institutional trust and existing contracts. What OpenAI has is the model capability advantage and an increasingly credible argument that it can run the intelligence layer across the whole enterprise, not just one product family.
What CIOs are actually deciding
Frontier is currently available to a limited set of customers, with broader availability expected over the coming months. Pricing has not been disclosed publicly, with OpenAI directing interested organisations to its enterprise sales team.
For CIOs, the practical decision is not yet binary. Most large enterprises run Salesforce, ServiceNow, and Microsoft infrastructure simultaneously. The immediate question is whether Frontier becomes an orchestration layer that connects those systems, or a competitive platform that starts displacing them.
OpenAI’s chief revenue officer, Denise Dresser, offered what is probably the most honest summary of where enterprise AI agents stand right now. “What’s really missing still for most companies is just a simple way to unleash the power of agents as teammates that can operate inside the business without the need to rework everything underneath.”
That gap is exactly what every platform in this space claims to close. The difference with Frontier is that the company making the claim now has the enterprise relationships, the production deployments, and the model capability to back it up. The SaaS incumbents have a head start on trust and data. Whether that proves sufficient is the central question for enterprise software through the rest of 2026.
(Photo by Austin Distel)
See also: OpenAI’s enterprise push: The hidden story behind AI’s sales race
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post OpenAI Frontier puts enterprise AI agents at the centre of a fight the SaaS industry cannot afford to lose appeared first on AI News.
View the full article
E.SUN Bank is working with IBM to build clearer AI governance rules for how artificial intelligence can be used inside a bank. The effort reflects a wider shift in finance. Many firms already use AI for fraud checks and credit scoring, and some also use it to handle customer service queries. The new challenge is how to manage these systems in a way that meets legal and risk rules.
Banks face a growing list of questions as they deploy AI. How should a model be tested before it goes live? Who is responsible if it makes a wrong call? And how can firms prove to regulators that their systems are fair and safe?
To address those issues, E.SUN Bank and IBM Consulting have created an AI governance framework for banking. The project also includes an AI governance white paper that sets out how financial firms can build internal controls around AI systems. According to the companies’ press release, the work adapts global standards such as the EU AI Act and ISO/IEC 42001 for financial services.
The framework sets out how banks can review AI models before they are deployed. It also explains how those models should be monitored after they enter production. It includes rules for how data is used and how risk reviews should take place.
E.SUN Bank said the framework is intended to help financial institutions introduce AI systems while maintaining governance and regulatory oversight. Many firms already run limited AI tools. The next step is to scale those systems across core operations such as lending and payments while staying within regulatory limits.
Banks try to manage AI risk
Financial firms have strong reasons to place guardrails around AI systems. Banking relies on trust, and regulators require firms to track how decisions are made. AI models often act as “****** boxes,” meaning it can be hard to explain how they arrive at a result. That can create problems in areas such as credit decisions or fraud checks. Regulators in many regions have started to focus on these risks.
The European Union’s AI Act, adopted in 2024, places strict rules on AI systems used in high-risk sectors such as finance. The law requires firms to assess risks and document training data. It also requires them to monitor how AI models behave after deployment.
Global standards are also taking shape. ISO/IEC 42001, published in 2023, sets out how organisations can build management systems for AI. The standard focuses on oversight and model monitoring. It also addresses how organisations should manage AI data. The aim is to give firms a structured way to manage AI across an entire company rather than treating each model as a separate tool.
E.SUN Bank’s project with IBM draws from both frameworks. It is meant to show how these rules could work in daily banking operations.
From AI pilots to enterprise systems
Banks have used machine learning for years, mainly in risk analysis and fraud detection. Newer AI models are expanding how banks use the technology. Many now apply it in customer service and document review. Some also use it in internal knowledge systems.
That expansion brings new governance needs. A system that suggests answers to customer queries may seem low risk. But a model that helps approve loans or detect fraud can have direct financial effects.
The governance framework created by E.SUN Bank and IBM sets out a process to track those risks. Models are reviewed before they go live, and teams monitor their output after deployment. The framework also assigns responsibility across teams, from developers to compliance staff. The project also produced a white paper that explains the steps in more detail. It outlines how banks can classify AI systems by risk level and apply different levels of oversight.
AI governance expands across financial services
The work at E.SUN Bank reflects a trend across global finance. Many banks now see governance as a key step before scaling AI across operations.
Industry surveys suggest that AI adoption in financial services is already widespread. A 2024 report by NVIDIA found that about 91% of financial services firms were either assessing or already using AI. Common uses include fraud detection and risk modelling. Some banks also use AI to automate customer service tasks.
Research from Deloitte shows that more than 70% of financial institutions plan to increase investment in AI. Much of that spending is aimed at compliance monitoring and risk analysis. Some banks also expect AI to improve internal operations.
At the same time, regulators are paying closer attention. Authorities in several regions have warned banks to track how automated systems affect decisions such as credit approval and fraud detection. This pressure has led banks to invest more in internal oversight systems. Instead of focusing only on model accuracy, firms now also track data sources and decision logic. Many also monitor how models behave over time.
Why governance may shape AI adoption
The push for AI governance may influence how quickly banks adopt new tools. Without clear rules, many firms hesitate to move beyond small experiments. A structured framework can help them expand AI projects while still meeting regulatory demands.
That is the idea behind the E.SUN Bank project. By combining global standards with banking workflows, the framework sets out how AI can be deployed under clear oversight. According to the companies’ announcement, IBM said the framework was developed to help financial institutions manage AI risks as they expand their use of AI in banking.
The effort also reflects the growing role of governance in enterprise AI. Early AI projects focused on building models and improving performance. Today the focus is shifting toward how those systems are managed over time. As more banks bring AI into core operations, that question may become just as important as the technology itself.
(Photo by Markus Spiske)
See also: Manulife moves AI agents into core financial workflows
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post E.SUN Bank and IBM build AI governance framework for banking appeared first on AI News.
View the full article
Europe’s factory floors have a new kind of colleague. BMW Group has deployed humanoid robots in manufacturing in Germany for the first time, launching a pilot project at its Leipzig plant with AEON–a wheeled humanoid built by Hexagon Robotics.
It is the first automotive deployment of AEON anywhere in the world, and it marks something of a line in the sand for European industry: physical AI is no longer a North American or East Asian story.
The announcement, made on March 9, 2026, comes backed by hard data from a prior US trial. In 2025, BMW ran a ten-month pilot at its Spartanburg, South Carolina, plant using Figure AI’s Figure 02 robot. The humanoid supported production of over 30,000 BMW X3s, working 10-hour shifts and moving a total of over 90,000 components.
Leipzig is now the direct heir to those lessons.
A robot built for work, not demos
AEON, developed by Hexagon’s Zurich-based robotics division, is a deliberately industrial machine. Arnaud Robert, President of Hexagon Robotics, made the philosophy plain at a Munich event earlier this month: “We’re not in the dancing business–we’re in the working business.” That ethos is visible in every design decision.
Rather than walking on two legs, AEON moves on wheels–a choice made after extensive testing of locomotion systems, with Hexagon concluding that on factory-grade flat floors, wheels are significantly more efficient in both speed and energy use. It stands 1.65 metres tall, weighs 60 kilograms, reaches 2.5 metres per second, and can autonomously swap its own battery in 23 seconds–enabling around-the-clock operation without human intervention.
Its 22 integrated sensors–peripheral cameras, time-of-flight, infrared, SLAM cameras, and microphones–give it full 360-degree real-time spatial awareness, including the ability to perform quality inspection tasks that conventional stationary robots cannot.
Its human-like torso allows a wide variety of grippers, hand elements, and scanning tools to be flexibly docked, which is precisely what BMW needs for multifunctional deployment across different production environments
Phased rollout, deliberate strategy
AEON’s first test deployment at Leipzig took place in December 2025. A further test run is planned for April 2026, ahead of a full pilot phase launching in summer 2026, where two AEON units will work simultaneously across two use cases–focusing on high-voltage battery assembly and component manufacturing for exterior parts.
Leipzig was not an arbitrary choice. It is BMW’s most technologically comprehensive ******* plant, combining battery production, injection moulding, press shop, body shop, and final assembly under one roof, meaning a successful deployment there effectively validates physical AI across the full production spectrum.
To anchor this work institutionally, BMW has established a Centre of Competence for Physical AI in Production, consolidating expertise across the group and creating a defined evaluation path for technology partners–from lab testing through to full pilot phases.
As Felix Haeckel, Team Lead for the centre, put it: “We are pooling our expertise to make knowledge on AI and robotics widely usable within the company.”
The infrastructure underneath
What makes BMW’s approach notable is that AEON is not landing on a blank factory floor. BMW has systematically dismantled data silos across its production network, replacing them with a uniform data platform that ensures all information is consistent, standardised, and accessible at all times–the architecture that allows AI agents to operate autonomously and learn continuously.
The humanoid robot is, in effect, the physical layer of a system that has been years in the making. AEON runs on NVIDIA Jetson Orin onboard computers and was trained largely through simulation using NVIDIA’s Isaac platform–a method that allowed Hexagon to develop core locomotion capabilities in weeks rather than months.
The project also involves Microsoft Azure for scalable model development and Maxon’s actuators for locomotion.
Why this matters beyond Leipzig
The broader signal here is one that the enterprise AI world is already tracking closely. Deloitte’s State of AI in the Enterprise 2026 report, surveying over 3,200 senior leaders across 24 countries, found that 58% of companies are already using physical AI in some capacity, with that figure set to reach 80% within two years, with Asia Pacific leading in early implementation.
BMW’s Leipzig pilot is a proof point in that trajectory: that humanoid robots in manufacturing have moved past the lab and the press release, and are being stress-tested against the unforgiving standards of real industrial production. As Milan Nedeljković, BMW’s Board Member for Production, put it: “The symbiosis of engineering expertise and artificial intelligence opens up completely new possibilities in production.”
The question now is not whether humanoid robots belong on the factory floor. It is how fast the rest of the European industry follows.
See also: Ai2: Building physical AI with virtual simulation data
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post BMW puts humanoid robots to work in Germany–and Europe’s factories are watching appeared first on AI News.
View the full article
Privacy Notice: We utilize cookies to optimize your browsing experience and analyze website traffic. By consenting, you acknowledge and agree to our Cookie Policy, ensuring your privacy preferences are respected.
