San Francisco: Microsoft on Tuesday announced plans to unveil a more open and inclusive Windows operating system for developers, seven new artificial intelligence (AI) models, and an operating system (OS) for AI-first devices. The announcements came on the same day that US President Donald Trump signed an executive order, which gave the government power to scrutinize AI models in the interest of “national security”.
The company put up a bullish stance despite the US government’s decision, which will require all tech companies to give the government access to frontier AI models 30 days before releasing them commercially.
Microsoft “has a long-standing history of partnering with governments around the world, and being very thoughtful in creating policies that have always respected regulatory fragmentations of every sovereign nation,” Dave Citron, corporate vice-president of Microsoft AI, told Mint.
In a keynote spanning over two hours, chairman and chief executive Satya Nadella spoke about one of the most significant updates to Microsoft’s Windows operating system, which has over 1.4 billion users worldwide and over 100 million users in India.
The updates include new, small AI models that run locally on Windows laptops, new tools to make Windows more compatible for AI developers working across platforms, including Apple’s Macs, new PCs in partnership with Nvidia, and upcoming custom processors that Nadella claimed will offer “the cheapest performance per dollar per watt.”
Project Solara, for which Microsoft is working with both Qualcomm and Mediatek–the world’s two largest mobile chipmakers, appears to be building an agentic AI platform which adjusts its interface to a wide range of devices, “including form factors and devices that do not exist today,” said Steven Bathiche, corporate vice-president and technical fellow at Microsoft.
Mustafa Suleyman, the company’s AI chief, launched its new AI models, including the company’s first 35-billion-parameter “thinking” model, which the company claimed is on a par with or better than models from key rivals Anthropic, Google and OpenAI.
Citron told Mint that the company will continue supporting its partnerships with Anthropic and OpenAI as well, alongside offering enterprise clients the option to use its in-house models.
The company also unveiled M-Dash (multi-model agentic scanning harness)—a cyber security platform that was developed alongside the company’s partnership with Anthropic’s much-debated Claude Mythos and Project Glasswing. The security platform will autonomously scan technical infrastructure for critical risks and also patch them. M-Dash is debuting globally, including in India, with Accenture and PricewaterhouseCoopers as two of its first customers, Vasu Jakkal, corporate vice-president of Microsoft Security told Mint.
“We want Windows to be a fantastic place to run and scale AI agents, and we’re building everything that developers need to make AI work for frontier firms. This includes fully autonomous agents… they work where people work. We’re excited to be doing all of this, as well as a new platform for the AI era that goes beyond just one form factor,” Nadella said.
News of President Trump’s executive order dampenened sentiment for tech stocks in the US. After the announcements, Microsoft’s share price dropped 4.17% on Tuesday, and a further 0.2% in after-hours trading. Shares of other AI-first tech giants – Amazon, Google and Nvidia – declined as well.
Industry analysts said the announcements were largely in line with expectations.
Jim Mercer, program vice-president at independent consultancy and research firm International Data Corporation (IDC), said Microsoft’s announcements “were largely on-theme, given the current balance of circumstances in the technology industry.”
“The key thing for Microsoft is to be able to win back developer mind-share from new firms such as Anthropic and OpenAI, which are going aggressive on innovation and almost have nothing much to lose—as compared with Microsoft that has been around for decades,” said Mercer, who was also at Microsoft’s annual developer conference.
“The company’s all-out AI innovation approach is somewhat more conservative than some of its newest rivals, but it’s important to note that Nadella’s leadership is proven and sharp—and the focus on developers and Windows is with a clear eye on the long run.”
Anushree Verma, senior director analyst at Gartner, said the company is “carefully building out completely independent capabilities that give it full control of the entire AI stack.”
“Microsoft’s OpenAI partnership, though fruitful early on, hasn’t quite paid long-term dividends as the two companies chase different paths of progress. This makes Project Solara and its foundational models crucial for the long run, and Microsoft is one of the few around the world–alongside Google–to have access to enough capital to chase both market-facing applications and moonshot innovations alike,” Verma said.
The writer is in San Francisco on Microsoft’s invitation.
