As synthetic intelligence (AI) continues to remodel industries throughout Europe and past, Microsoft has unveiled improvements that promise to reshape every thing from cloud sovereignty to environmental forecasting. This is a more in-depth take a look at 4 main developments from Microsoft and its collaborators.
Additionally Learn: Microsoft to Improve European Information Middle Capability by 40 P.c, Proclaims 5 Digital Commitments
1. Microsoft Expands Sovereign Cloud Choices to Strengthen Digital Sovereignty in Europe
Microsoft has unveiled an growth of its Sovereign Cloud portfolio, reinforcing its dedication to digital sovereignty, information privateness, and operational resilience for European clients. In line with Microsoft, the announcement builds on its 42-year presence within the area and aligns with its European Digital Commitments to supply higher management, compliance, and option to governments and controlled industries.
“With Microsoft Sovereign Cloud, we’re dedicated to providing essentially the most complete set of sovereignty options for purchasers throughout Sovereign Public Cloud, Sovereign Non-public Cloud, and the Nationwide Companion Clouds operated by our European companions,” Microsoft CEO Satya Nadella introduced on platform X on Monday, June 16, 2025.
“At present, we’re saying new choices that convey digital sovereignty to all European organizations within the public cloud and unlock new methods to run non-public sovereign clouds too:
Relating to the general public cloud, we’re introducing Information Guardian, which ensures all distant entry by our engineers to the methods that retailer and course of buyer information in Europe is accepted and monitored by European resident personnel in real-time and logged in a tamper-evident ledger.
In addition to Exterior Key Administration for customer-controlled encryption, offering an extra assure of knowledge safety.
We’re additionally introducing Sovereign Non-public Cloud, which incorporates Microsoft 365 Native, which brings collectively our productiveness server software program into an Azure Native setting that may run solely in a buyer’s personal datacenter, with full management on safety, compliance and governance.
These new choices construct on a long time of pioneering work in sovereign cloud options by ourselves and our companions, and replicate our ongoing dedication to providing you with extra selection, management, and safety,” Nadella introduced.
The expanded choices embrace new capabilities inside Microsoft’s Sovereign Public Cloud, the introduction of a Sovereign Non-public Cloud with Azure Native, and enhanced Nationwide Companion Cloud preparations in France and Germany. These options are designed to offer European clients with the pliability to run workloads in both public or non-public environments, making certain full alignment with European legal guidelines and information residency necessities.
Among the many key options introduced is Information Guardian, a service that enforces European-only personnel management over distant entry to buyer information, supported by real-time oversight and tamper-evident logging. Exterior Key Administration, one other addition, permits organizations to make use of their very own {Hardware} Safety Modules (HSMs)—both on-premises or hosted by trusted third events—for full encryption management. Microsoft is working with main HSM suppliers together with Futurex, Thales, and Utimaco to help the combination.
The corporate additionally launched Regulated Surroundings Administration, a unified interface for managing sovereign configurations, insurance policies, and audit logs. This service simplifies the deployment and governance of delicate workloads in compliance with nationwide and sector-specific rules.
Microsoft’s Sovereign Public Cloud, an evolution of its Cloud for Sovereignty, will now be out there throughout all European datacenter areas. It permits enterprise clients to retain management over their information with out requiring migration to separate environments, whereas nonetheless benefiting from the capabilities of Microsoft Azure, Microsoft 365, Microsoft Safety, and Energy Platform.
In parallel, Microsoft has introduced launching a Sovereign Non-public Cloud, at present in preview, to help workloads that should stay in totally customer-controlled environments. Constructed on Azure Native infrastructure, it permits deployment of key Microsoft cloud companies, together with compute and storage, inside on-premises or in-country datacenters. Complementing this, Microsoft 365 Native brings acquainted productiveness purposes like Alternate and SharePoint to customer-operated environments, providing full management over information, safety, and governance.
In France and Germany, Microsoft continues to companion with nationwide entities to ship sovereign cloud capabilities. In France, Bleu—a three way partnership between Orange and Capgemini—operates a “cloud de confiance” for the general public sector and important infrastructure suppliers, compliant with SecNumCloud necessities. In Germany, Delos Cloud, a subsidiary of SAP, manages a sovereign cloud resolution tailor-made to the German authorities’s platform wants.
To help clients in implementing these options, Microsoft can be previewing a Sovereign Cloud specialization inside its AI Cloud Companion Program. This initiative highlights companions with verified capabilities in deploying sovereign workloads. Preview companions embrace Accenture, Arvato Programs, Atea, Atos, Crayon, Capgemini, Dell Applied sciences, IBM, Inspark, Infosys, Lenovo, Leonardo, NTT Information, Orange, Telefonica and Vodafone.
“The launch of Microsoft Sovereign Cloud marks a pivotal second in empowering European establishments and industries with the management, compliance and innovation they should thrive in at present’s digital financial system,” stated Aiman Ezzat, CEO of Capgemini Group.
“As a shareholder of Bleu, we’ve got already arrange a Nationwide Companion Cloud in France with a purpose to ship Microsoft applied sciences in a sovereign setting that respects the French State necessities. With a long time of expertise in Microsoft applied sciences and deep experience in regulated sectors, we’re uniquely positioned to assist our purchasers harness the complete energy of Microsoft’s sovereign private and non-private cloud options. Collectively, we’re enabling a trusted digital future for Europe.”
Microsoft’s Sovereign Cloud portfolio, the corporate says, now represents essentially the most complete suite of sovereignty-focused options out there. It combines productiveness, safety, and cloud infrastructure with built-in compliance and transparency, aiming to make sure that European clients can function with confidence in an more and more advanced regulatory and geopolitical setting.
Additionally Learn: Europe’s Quickest AI Supercomputer, Jupiter, Launched with Nvidia Grace Hopper Platform
2. Albert Heijn’s AI Assistant Steijn Transforms Grocery Purchasing and Meal Planning for Prospects in Netherlands
Dutch grocery store large Albert Heijn has launched an AI-powered assistant named Steijn to assist clients make smarter, more healthy, and extra environment friendly meals decisions. Built-in into the Albert Heijn cellular app, the assistant is already altering the way in which consumers plan meals, cut back meals waste, and store for groceries, based on Microsoft.
“You already know AI is transformative when it’s at your dinner desk. Within the Netherlands at present, I met the Albert Heijn crew who’re utilizing Azure AI Foundry to assist clients navigate on a regular basis choices, like what’s for dinner,” Microsoft CEO Satya Nadella shared in a publish on X on June 16, 2025.
Steijn is constructed utilizing Azure AI Foundry and powered by OpenAI fashions, providing clients real-time recipe options based mostly on their preferences, dietary wants, and even the contents of their fridge. The software attracts from a database of greater than 20,000 recipes from Allerhande (Every kind), the model’s meals journal, and offers dietary insights, preparation ideas, and personalised suggestions.
Sjoerd Holleman, Senior Vice President for Technique, Product and Analytics at Albert Heijn, emphasised that the assistant was designed to reply the on a regular basis query: “What’s for dinner tonight?” Whether or not customers are searching for vegetarian, gluten-free, or 20-minute recipes, Steijn can rapidly generate appropriate choices tailor-made to particular wants.
Steijn additionally contains picture recognition options. Customers can take a photograph of the within of their fridge, and the assistant will recommend meals utilizing the elements out there. This performance has confirmed notably helpful for purchasers like Sander van Straaten, a advertising freelancer and father of two in Amsterdam. It saves me time and helps cut back meals waste, he stated. It even remembers my preferences for fast, wholesome meals.
Developed in three months by an eight-member crew led by product supervisor Norman van Ameyden, Steijn is without doubt one of the first large-scale customer-facing purposes developed with Azure AI Foundry within the Netherlands.
“The instruments of Microsoft gave us the chance to essentially construct Steijn with a small crew in a brief period of time,” van Ameyden stated.
He stated his crew examined completely different prospects utilizing the chat playground in Azure AI Foundry portal whereas constructing Steijn, which is powered by OpenAI in Foundry Fashions and Azure AI Search. Azure AI Foundry is a platform designed to assist builders and firms design, customise and handle AI purposes and brokers with relative ease at a big scale. Playground is a testing space inside Foundry.
“We wished to verify we are able to do fast POCs (proofs of idea) and experiments, and that actually helped us, as a result of we did a number of testing simply earlier than we went to manufacturing,” he stated. “And Azure AI Foundry was excellent for that.”
Information privateness and security are a key focus. Conversations with Steijn are anonymized and deleted each 30 days. Moreover, Microsoft’s Azure AI Content material Security instruments are used to observe and filter any dangerous content material.
Albert Heijn, which operates 1,200 shops throughout the Netherlands and northern Belgium, serves roughly 5 million clients weekly. The app itself additionally helps self-scanning and personalised reductions, whereas superior forecasting algorithms make as much as one billion predictions a day to optimize inventory and pricing. Digital shelf labels in shops alter reductions routinely as merchandise close to their expiration date, serving to additional cut back waste.
Trying forward, Albert Heijn plans to broaden Steijn’s capabilities to additional help its sustainability targets, together with a 50 p.c discount in meals waste throughout your complete meals chain by 2030.
Additionally Learn: Microsoft Proclaims USD 400 Million Funding to Enhance Cloud and AI Infrastructure in Switzerland
3. Barclays to Roll Out Microsoft 365 Copilot to 100,000 Staff Globally
Barclays Financial institution (Barclays) and Microsoft introduced that Barclays is to roll out Microsoft 365 Copilot to 100,000 colleagues globally, remodeling the worker expertise by AI brokers and giving them entry to the knowledge they want most.
The deployment will combine Microsoft 365 Copilot into Barclays’ proprietary colleague productiveness software, making a single agent that allows staff to entry the financial institution’s ecosystem of collaboration instruments, portals and on-line assets all from one place, making it less complicated to seek out data, enhance productiveness, and improve worker expertise, Microsoft stated on June 9, 2025.
“Barclays is scaling Microsoft 365 Copilot to 100,000 staff, placing AI in each worker’s fingers. This may simplify how they entry data, get issues completed, and make Copilot the UI for Barclays AI,” Satya Nadella shared in a publish on X on June 9, 2025.
As a part of the rollout, Barclays staff will acquire entry to a brand new AI-powered Colleague Agent by Microsoft 365 Copilot. This software will present self-service capabilities built-in with Barclays’ inside methods and third-party connectors. Staff will be capable of full a variety of duties—similar to reserving journey, checking coverage compliance, or discovering HR-related data—with ease and effectivity.
A brand new semantic content material search function will additional help productiveness by delivering related, personalised outcomes based mostly on consumer profiles and site. As well as, a dashboard generally known as the “Colleague Entrance Door,” constructed on Microsoft Viva, will function a central hub for high-priority actions like desk bookings or annual go away requests, whereas additionally serving personalised information and bulletins.
Commenting on the settlement, Craig Shiny, Group Chief Data Officer and Deputy Group Co-Chief Working Officer, Barclays, stated: “At Barclays, we’ve been leveraging the ability of AI, and now GenAI, to drive deeper insights, enhance effectivity and create extra intuitive expertise throughout the organisation.
“Our roll-out of Copilot, built-in with our colleague productiveness software, is a major step ahead in simplifying the way in which we work, making it simpler to get issues completed.
“It additionally highlights the collaborative partnership with Microsoft – the place innovation is formed by sensible utility at scale.”
The announcement follows a profitable pilot involving 15,000 Barclays staff and builds upon the financial institution’s adoption of Microsoft Groups for collaboration and Microsoft Viva Interact for worker engagement.
Darren Hardman, CEO, Microsoft UK, added: “The adoption of Microsoft 365 Copilot to be the UI for Barclays AI will assist them to ship on their daring imaginative and prescient of placing AI within the fingers of each worker, and we stay up for working intently with Barclays to assist its colleagues maximise the advantages from utilizing this transformational know-how.”
Additionally Learn: Microsoft to Make investments USD 3 Billion in India to Enhance AI, Cloud, and Skilling: CEO
4. Microsoft Says AI Mannequin Aurora Predicts Excessive Climate with Extra Accuracy
As local weather change intensifies the frequency and severity of maximum climate occasions, Microsoft has unveiled an AI mannequin able to remodeling international environmental forecasting. In a brand new paper printed in Nature, Microsoft Analysis launched Aurora, a basis mannequin skilled on the most important assortment of atmospheric information, able to predicting not solely climate patterns however a wide selection of environmental phenomena — together with hurricanes, air air pollution, and ocean wave exercise.
“Developed by Microsoft Analysis, Aurora additionally forecasts this vary of atmospheric occasions with higher precision and velocity and at a lot decrease computational price when in comparison with conventional numerical forecasting and former AI approaches,” Microsoft stated in a weblog publish, noting that the muse mannequin leverages the most recent advances in AI to extra precisely predict climate and a variety of environmental occasions.
“Aurora is a brand new basis mannequin from Microsoft Analysis that goes past climate forecasting, delivering sooner, extra correct predictions of environmental occasions. Superior to see this breakthrough printed in Nature,” shared Satya Nadella in a publish on X on Could 21, 2025.
“Aurora first learns tips on how to generate forecasts in simply seconds by coaching on normal climate patterns from over a million hours of knowledge derived from satellites, radar and climate stations, simulations and forecasts. Microsoft researchers imagine it’s the largest assortment of atmospheric information ever assembled to coach an AI forecasting mannequin. Profiting from its distinctive versatile structure, Aurora is then ‘fine-turned’ to carry out a wide range of particular duties similar to predicting wave top or air high quality, utilizing modest quantities of further information,” Microsoft defined.
In contrast to conventional AI fashions restricted to particular duties, Aurora leverages a foundational structure that permits for high-performance generalization throughout a number of forecasting challenges. Aurora delivers medium-range climate forecasts — as much as 14 days — with accuracy, velocity, and computational effectivity.
In line with Microsoft, Aurora outperforms standard numerical climate fashions and present AI methods throughout 91 p.c of forecasting targets when fine-tuned to a decision of 0.25 levels. Crucially, it additionally delivers forecasts as much as 5,000 occasions sooner than conventional fashions, producing leads to seconds utilizing high-bandwidth GPUs, whereas working at a considerably decrease price as soon as skilled.
Though the preliminary coaching of Aurora is dear, its operational bills are considerably decrease than these of conventional climate prediction methods as soon as it’s totally purposeful, the authors wrote in Nature.
Predicting the Unpredictable
In line with Microsoft, Aurora’s capability to anticipate excessive climate occasions has already proven tangible promise. In a retrospective evaluation, Aurora precisely predicted Hurricane Doksuri’s landfall within the Philippines 4 days earlier than the occasion, outperforming official forecasts that anticipated it to bypass the area.
On this newest analysis, Aurora additionally beat the US Nationwide Hurricane Middle in forecasting 5-day tropical cyclone tracks, a primary for a machine-learning mannequin. In a significant achievement, it additionally outperformed seven main forecasting facilities on all cyclone observe forecasts for the 2022-2023 season globally, the researchers discovered, Microsoft stated.
Past storms, Aurora demonstrated highly effective air high quality forecasting capabilities. It efficiently anticipated a extreme sandstorm in Iraq in June 2022, which grounded flights and despatched hundreds to hospitals. Regardless of restricted air air pollution information, Aurora’s switch studying capability — absorbing patterns from its foundational coaching — enabled correct fine-tuning with comparatively small datasets.
Ocean Insights and Environmental Versatility
Aurora additionally excelled in predicting ocean wave patterns. In assessments, it outperformed present fashions in 86 p.c of comparisons over a yr, even when skilled on just some years of knowledge. In a single case, it precisely projected the wave impacts of Hurricane Nanmadol, Japan’s most intense storm of 2022.
Microsoft’s researchers credit score Aurora’s success to its versatile structure and deep-learning method, which avoids hardcoding bodily guidelines. As a substitute, the mannequin learns patterns and interactions instantly from the information, permitting it to adapt to duties starting from precipitation monitoring to crop logistics and energy grid safety.
Open Entry and Sensible Deployment
To encourage international collaboration and innovation, Microsoft has made Aurora’s supply code and mannequin weights freely out there. The mannequin can be featured on Azure AI Foundry Labs and is built-in into MSN Climate, the place it enhances hourly forecasts with extra correct information on precipitation and cloud cowl.
Aurora’s modular design permits researchers and organizations to fine-tune the mannequin to their particular wants in a matter of weeks — a drastic distinction to conventional forecasting methods, which may take years to develop. Additionally it is accessible by the European Centre for Medium-Vary Climate Forecasts (ECMWF), one of many world’s broadly used meteorological companies, Microsoft stated.
Trying Forward
Microsoft stated Aurora’s early outcomes have stirred curiosity throughout academia and trade, together with forecasting organizations, power firms and even commodity merchants. There’s specific curiosity in seeing how it may be tailored to raised predict rain, improve crop logistics and shield power grids.
Whereas Microsoft emphasizes that Aurora isn’t meant to switch present climate methods, it sees the mannequin as a strong complement — particularly in underserved areas missing superior forecasting instruments.
Megan Stanley, a senior researcher at Microsoft, believes Aurora—and what comes subsequent—will complement present forecasting methods, not change them. “There’s a number of fascinating analysis to be completed round how effectively it’s studying the physics, and whether it is studying the physics appropriately then it means that is one thing that ought to be strong sufficient to make predictions in numerous climatic settings. It is the primary of its variety,” provides Stanley. “However it doesn’t imply it will likely be the final.”
You can too be a part of the TelecomTalk WhatsApp Neighborhood, TelecomTalk WhatsApp Channel or Telegram group Telecom Circle for updates and discussions.
