Microsoft Expands Sovereign Cloud Offerings in Europe, AI Assistant Steijn Meal Planning, and More

As well as External Key Management for customer-controlled encryption, providing an additional guarantee of data protection.

We’re also introducing Sovereign Private Cloud, which includes Microsoft 365 Local, which brings together our productivity server software into an Azure Local environment that can run entirely in a customer’s own datacenter, with full control on security, compliance and governance.

These new offerings build on decades of pioneering work in sovereign cloud solutions by ourselves and our partners, and reflect our ongoing commitment to giving you more choice, control, and security,” Nadella announced.

The expanded offerings include new capabilities within Microsoft’s Sovereign Public Cloud, the introduction of a Sovereign Private Cloud with Azure Local, and enhanced National Partner Cloud arrangements in France and Germany. These solutions are designed to provide European customers with the flexibility to run workloads in either public or private environments, ensuring full alignment with European laws and data residency requirements.

With Microsoft’s Sovereign Public Cloud currently in preview and set to be generally available in all European cloud regions later this year, Microsoft said it will introduce new features and solutions that reinforce this vision.

Among the key features announced is Data Guardian, a service that enforces European-only personnel control over remote access to customer data, supported by real-time oversight and tamper-evident logging. External Key Management, another addition, enables organizations to use their own Hardware Security Modules (HSMs)—either on-premises or hosted by trusted third parties—for full encryption control. Microsoft is working with leading HSM providers including Futurex, Thales, and Utimaco to support the integration.

The company also introduced Regulated Environment Management, a unified interface for managing sovereign configurations, policies, and audit logs. This service simplifies the deployment and governance of sensitive workloads in compliance with national and sector-specific regulations.

Microsoft’s Sovereign Public Cloud, an evolution of its Cloud for Sovereignty, will now be available across all European datacenter regions. It allows enterprise customers to retain control over their data without requiring migration to separate environments, while still benefiting from the capabilities of Microsoft Azure, Microsoft 365, Microsoft Security, and Power Platform.

In parallel, Microsoft has announced launching a Sovereign Private Cloud, currently in preview, to support workloads that must remain in fully customer-controlled environments. Built on Azure Local infrastructure, it enables deployment of key Microsoft cloud services, including compute and storage, within on-premises or in-country datacenters. Complementing this, Microsoft 365 Local brings familiar productivity applications like Exchange and SharePoint to customer-operated environments, offering full control over data, security, and governance.

In France and Germany, Microsoft continues to partner with national entities to deliver sovereign cloud capabilities. In France, Bleu—a joint venture between Orange and Capgemini—operates a “cloud de confiance” for the public sector and critical infrastructure providers, compliant with SecNumCloud requirements. In Germany, Delos Cloud, a subsidiary of SAP, manages a sovereign cloud solution tailored to the German government’s platform needs.

To support customers in implementing these solutions, Microsoft is also previewing a Sovereign Cloud specialization within its AI Cloud Partner Program. This initiative highlights partners with verified capabilities in deploying sovereign workloads. Preview partners include Accenture, Arvato Systems, Atea, Atos, Crayon, Capgemini, Dell Technologies, IBM, Inspark, Infosys, Lenovo, Leonardo, NTT Data, Orange, Telefonica and Vodafone.

“The launch of Microsoft Sovereign Cloud marks a pivotal moment in empowering European institutions and industries with the control, compliance and innovation they need to thrive in today’s digital economy,” said Aiman Ezzat, CEO of Capgemini Group.

“As a shareholder of Bleu, we have already set up a National Partner Cloud in France in order to deliver Microsoft technologies in a sovereign environment that respects the French State requirements. With decades of experience in Microsoft technologies and deep expertise in regulated sectors, we are uniquely positioned to help our clients harness the full power of Microsoft’s sovereign public and private cloud solutions. Together, we are enabling a trusted digital future for Europe.”

Microsoft’s Sovereign Cloud portfolio, the company says, now represents the most comprehensive suite of sovereignty-focused solutions in the market. It combines productivity, security, and cloud infrastructure with built-in compliance and transparency, aiming to ensure that European customers can operate with confidence in an increasingly complex regulatory and geopolitical environment.

Also Read: Europe’s Fastest AI Supercomputer, Jupiter, Launched with Nvidia Grace Hopper Platform

2. Albert Heijn’s AI Assistant Steijn Transforms Grocery Shopping and Meal Planning for Customers in Netherlands

Dutch supermarket giant Albert Heijn has introduced an AI-powered assistant named Steijn to help customers make smarter, healthier, and more efficient food choices. Integrated into the Albert Heijn mobile app, the assistant is already changing the way shoppers plan meals, reduce food waste, and shop for groceries, according to Microsoft.

“You know AI is transformative when it’s at your dinner table. In the Netherlands today, I met the Albert Heijn team who are using Azure AI Foundry to help customers navigate everyday decisions, like what’s for dinner,” Microsoft CEO Satya Nadella shared in a post on X on June 16, 2025.

Steijn is built using Azure AI Foundry and powered by OpenAI models, offering customers real-time recipe suggestions based on their preferences, dietary needs, or even the contents of their refrigerator. The tool draws from a database of more than 20,000 recipes from Allerhande (All kinds), the brand’s food magazine, and provides nutritional insights, preparation tips, and personalized recommendations.

Sjoerd Holleman, Senior Vice President for Strategy, Product and Analytics at Albert Heijn, emphasized that the assistant was designed to answer the everyday question: “What’s for dinner tonight?” Whether users are looking for vegetarian, gluten-free, or 20-minute recipes, Steijn can quickly generate suitable options tailored to specific needs.

Steijn also includes image recognition features. Users can take a photo of the inside of their refrigerator, and the assistant will suggest meals using the ingredients available. This functionality has proven particularly useful for customers like Sander van Straaten, a marketing freelancer and father of two in Amsterdam. It saves me time and helps reduce food waste, he said. It even remembers my preferences for quick, healthy meals.

Developed in three months by an eight-member team led by product manager Norman van Ameyden, Steijn is one of the first large-scale customer-facing applications developed with Azure AI Foundry in the Netherlands.

“The tools of Microsoft gave us the opportunity to really build Steijn with a small team in a short amount of time,” van Ameyden said.

He said his team tested different possibilities using the chat playground in Azure AI Foundry portal while building Steijn, which is powered by OpenAI in Foundry Models and Azure AI Search. Azure AI Foundry is a platform designed to help developers and companies design, customize and manage AI applications and agents with relative ease at a large scale. Playground is a testing area within Foundry.

“We wanted to make sure we can do quick POCs (proofs of concept) and experiments, and that really helped us, because we did a lot of testing just before we went to production,” he said. “And Azure AI Foundry was perfect for that.”

Data privacy and safety are a key focus. Conversations with Steijn are anonymized and deleted every 30 days. Additionally, Microsoft’s Azure AI Content Safety tools are used to monitor and filter any harmful content.

Albert Heijn, which operates 1,200 stores across the Netherlands and northern Belgium, serves approximately five million customers weekly. The app itself also supports self-scanning and personalized discounts, while advanced forecasting algorithms make up to one billion predictions a day to optimize stock and pricing. Electronic shelf labels in stores adjust discounts automatically as products near their expiration date, helping further reduce waste.

Looking ahead, Albert Heijn plans to expand Steijn’s capabilities to further support its sustainability goals, including a 50 percent reduction in food waste across the entire food chain by 2030.

Also Read: Microsoft Announces USD 400 Million Investment to Boost Cloud and AI Infrastructure in Switzerland

3. Barclays to Roll Out Microsoft 365 Copilot to 100,000 Employees Globally

Barclays Bank (Barclays) and Microsoft announced that Barclays is to roll out Microsoft 365 Copilot to 100,000 colleagues globally, transforming the employee experience through AI agents and giving them access to the information they need most.

The deployment will integrate Microsoft 365 Copilot into Barclays’ proprietary colleague productivity tool, creating a single agent that enables employees to access the bank’s ecosystem of collaboration tools, portals and online resources all from one place, making it simpler to find information, improve productivity, and enhance employee experience, Microsoft said on June 9, 2025.

“Barclays is scaling Microsoft 365 Copilot to 100,000 employees, putting AI in every employee’s hands. This will simplify how they access information, get things done, and make Copilot the UI for Barclays AI,” Satya Nadella shared in a post on X on June 9, 2025.

As part of the rollout, Barclays employees will gain access to a new AI-powered Colleague Agent through Microsoft 365 Copilot. This tool will provide self-service capabilities integrated with Barclays’ internal systems and third-party connectors. Employees will be able to complete a range of tasks—such as booking travel, checking policy compliance, or finding HR-related information—with ease and efficiency.

A new semantic content search feature will further support productivity by delivering relevant, personalized results based on user profiles and location. In addition, a dashboard known as the “Colleague Front Door,” built on Microsoft Viva, will serve as a central hub for high-priority actions like desk bookings or annual leave requests, while also serving personalised news and announcements.

Commenting on the agreement, Craig Bright, Group Chief Information Officer and Deputy Group Co-Chief Operating Officer, Barclays, said: “At Barclays, we’ve been leveraging the power of AI, and now GenAI, to drive deeper insights, improve efficiency and create more intuitive experience across the organisation.

“Our roll-out of Copilot, integrated with our colleague productivity tool, is a significant step forward in simplifying the way we work, making it easier to get things done.

“It also highlights the collaborative partnership with Microsoft – where innovation is shaped by practical application at scale.”

The announcement follows a successful pilot involving 15,000 Barclays employees and builds upon the bank’s adoption of Microsoft Teams for collaboration and Microsoft Viva Engage for employee engagement.

Darren Hardman, CEO, Microsoft UK, added: “The adoption of Microsoft 365 Copilot to be the UI for Barclays AI will help them to deliver on their bold vision of putting AI in the hands of every employee, and we look forward to working closely with Barclays to help its colleagues maximise the benefits from using this transformational technology.”

Also Read: Microsoft to Invest USD 3 Billion in India to Boost AI, Cloud, and Skilling: CEO

4. Microsoft Says AI Model Aurora Predicts Extreme Weather with More Accuracy

As climate change intensifies the frequency and severity of extreme weather events, Microsoft has unveiled an AI model capable of transforming global environmental forecasting. In a new paper published in Nature, Microsoft Research introduced Aurora, a foundation model trained on the largest collection of atmospheric data, capable of predicting not only weather patterns but a wide array of environmental phenomena — including hurricanes, air pollution, and ocean wave activity.

“Developed by Microsoft Research, Aurora also forecasts this range of atmospheric events with greater precision and speed and at much lower computational cost when compared to traditional numerical forecasting and previous AI approaches,” Microsoft said in a blog post, noting that the foundation model leverages the latest advances in AI to more accurately predict weather and a range of environmental events.

“Aurora is a new foundation model from Microsoft Research that goes beyond weather forecasting, delivering faster, more accurate predictions of environmental events. Awesome to see this breakthrough published in Nature,” shared Satya Nadella in a post on X on May 21, 2025.

“Aurora first learns how to generate forecasts in just seconds through training on general weather patterns from over one million hours of data derived from satellites, radar and weather stations, simulations and forecasts. Microsoft researchers believe it is the largest collection of atmospheric data ever assembled to train an AI forecasting model. Taking advantage of its unique flexible architecture, Aurora is then ‘fine-turned’ to perform a variety of specific tasks such as predicting wave height or air quality, using modest amounts of additional data,” Microsoft explained.

Unlike traditional AI models limited to specific tasks, Aurora leverages a foundational architecture that allows for high-performance generalization across multiple forecasting challenges. Aurora delivers medium-range weather forecasts — up to 14 days — with accuracy, speed, and computational efficiency.

According to Microsoft, Aurora outperforms conventional numerical weather models and existing AI systems across 91 percent of forecasting targets when fine-tuned to a resolution of 0.25 degrees. Crucially, it also delivers forecasts up to 5,000 times faster than traditional models, generating results in seconds using high-bandwidth GPUs, while operating at a significantly lower cost once trained.

Although the initial training of Aurora is costly, its operational expenses are significantly lower than those of traditional weather prediction systems once it is fully functional, the authors wrote in Nature.

Predicting the Unpredictable

According to Microsoft, Aurora’s ability to anticipate extreme weather events has already shown tangible promise. In a retrospective analysis, Aurora accurately predicted Typhoon Doksuri’s landfall in the Philippines four days before the event, outperforming official forecasts that expected it to bypass the region.

In this latest research, Aurora also beat the US National Hurricane Center in forecasting 5-day tropical cyclone tracks, a first for a machine-learning model. In a major achievement, it also outperformed seven major forecasting centers on all cyclone track forecasts for the 2022-2023 season globally, the researchers found, Microsoft said.

Beyond storms, Aurora demonstrated powerful air quality forecasting capabilities. It successfully anticipated a severe sandstorm in Iraq in June 2022, which grounded flights and sent thousands to hospitals. Despite limited air pollution data, Aurora’s transfer learning ability — absorbing patterns from its foundational training — enabled accurate fine-tuning with relatively small datasets.

Ocean Insights and Environmental Versatility

Aurora also excelled in predicting ocean wave patterns. In tests, it outperformed current models in 86 percent of comparisons over a year, even when trained on just a few years of data. In one case, it accurately projected the wave impacts of Typhoon Nanmadol, Japan’s most intense storm of 2022.

Microsoft’s researchers credit Aurora’s success to its flexible architecture and deep-learning approach, which avoids hardcoding physical rules. Instead, the model learns patterns and interactions directly from the data, allowing it to adapt to tasks ranging from precipitation tracking to crop logistics and power grid protection.

Open Access and Practical Deployment

To encourage global collaboration and innovation, Microsoft has made Aurora’s source code and model weights freely available. The model is also featured on Azure AI Foundry Labs and is integrated into MSN Weather, where it enhances hourly forecasts with more accurate data on precipitation and cloud cover.

Aurora’s modular design allows researchers and organizations to fine-tune the model to their specific needs in a matter of weeks — a drastic contrast to traditional forecasting systems, which can take years to develop. It is also accessible through the European Centre for Medium-Range Weather Forecasts (ECMWF), one of the world’s widely used meteorological services, Microsoft said.

Looking Ahead

Microsoft said Aurora’s early results have stirred interest across academia and industry, including forecasting organizations, energy companies and even commodity traders. There is particular interest in seeing how it can be adapted to better predict rain, enhance crop logistics and protect energy grids.

While Microsoft emphasizes that Aurora is not intended to replace existing weather systems, it sees the model as a powerful complement — especially in underserved regions lacking advanced forecasting tools.

Megan Stanley, a senior researcher at Microsoft, believes Aurora—and what comes next—will complement current forecasting systems, not replace them. “There is a lot of interesting research to be done around how well it is learning the physics, and if it is learning the physics correctly then it means this is something that should be robust enough to make predictions in different climatic settings. It’s the first of its kind,” adds Stanley. “But it doesn’t mean it will be the last.”

You can also join the TelecomTalk WhatsApp Community, TelecomTalk WhatsApp Channel or Telegram group Telecom Circle for updates and discussions.