Expect this year to be filled with flux and uncertainty, thanks to lightning-fast changes wrought by AI, and by a Trump Administration that rules by whim and fiat rather than facts and law. So, I know it’s something of a fool’s game to make predictions about what the year ahead might bring for Microsoft.
That said, I’ve never worried about being called a fool. So here are my five predictions for what Microsoft can expect in 2026.
The AI bubble won’t burst — at least, not for Microsoft
There have been plenty of worries about what people believe is an AI bubble. As a McKinsey report cautions: “Nearly eight in 10 companies report using gen AI — yet just as many report no significant bottom-line impact.” Then there was that MIT report, The GenAI Divide: State of AI in Business 2025, which found that 95% of genAI pilots fail.
Despite that, Microsoft can’t keep up with demand for AI services. Its finance chief, Amy Hood, said during the company’s most recent earnings call, “I thought we were going to catch up [with demand]. We are not. Demand is increasing. It is not increasing in just one place. It is increasing across many places.”
She added that Microsoft has $400 billion under contract for future sales. “That’s for booked business,” she said. “Today.” (That number doesn’t include $250 billion in computing power for AI that OpenAI has agreed to buy from Microsoft.)
So while some AI companies will go under this year, Microsoft most likely won’t be affected. When it comes to AI, it’s full speed ahead for the company.
Moving to a new generation of AI
OpenAI laid the foundation for the generative AI (genAI) boom in November 2022 when it released ChatGPT, which eventually became the brains behind Microsoft’s Copilot. GenAI took off like a supercharged rocket, and that’s where most of the investment and publicity is at the moment. But it’s not at all clear there’s a big financial payoff from the technology.
Gary Marcus, a professor emeritus at New York University and founder of the AI startup Geometric Intelligence, argues in a New York Times opinion piece that genAI like ChatGPT and Copilot is not AI’s future, and may never pay off. He points to a report by Bain & Co that found there will be a $800 billion revenue shortfall for AI companies by the end of 2030 as evidence that another kind of AI is needed.
“If the strengths of AI are truly to be harnessed, the tech industry should stop focusing so heavily on these one-size-fits-all tools and instead concentrate on narrow, specialized AI tools engineered for particular problems,” Marcus wrote.
Microsoft believes the same thing. In late 2025, it laid out its vision of AI’s future, what it calls “Humanist Superintelligence.” Microsoft AI CEO and executive vice president Mustafa Suleyman argued that it will “solve real concrete problems and do it in such a way that it remains grounded and controllable. We are not building an ill-defined and ethereal superintelligence; we are building a practical technology explicitly designed only to serve humanity.”
Those are not just high-minded-sounding words — he said Microsoft has begun on what he calls Medical Superintelligence. Following that will be work on designing plentiful, clean, inexpensive energy.
Don’t be surprised if the first fruits of those labors start to show up this year.
Microsoft’s get-out-of-jail card
Apple, Amazon, Google, and Meta have all been hit with serious federal antitrust lawsuits during both the Biden and Trump administrations. But Microsoft, which was laid low by the mother of all antitrust suits back in 1998, has largely avoided them. In the one exception, the Federal Trade Commission (FTC) sued Microsoft because of the company’s purchase of Activision Blizzard. The agency lost. But even if it had won, the core of Microsoft’s business would have remained intact.
Still, Microsoft feared its core business would be targeted by the feds because in late November 2024, the FTC launched a broad antitrust investigation into the company’s AI, cloud computing, security, Teams, and AI products.
It’s been more than a year since that probe got under way, and there hasn’t been a peep about it. So it’s most likely dead. In addition, Trump has become Big AI’s biggest protector — he ordered the Justice Department to hit hard against any states that try to regulate AI.
So don’t expect him to target Microsoft for prosecution.
Copilot here, there and everywhere — for free
Although Microsoft has said its future AI plans will expand well beyond genAI, the company will likely push its genAI chatbot Copilot harder than ever in 2026. Microsoft says the chatbot will be included for free in all business versions of Microsoft 365 starting July 1. The price is currently up to $33.50 per user per month for businesses, depending on licensing arrangements and whether Microsoft is offering discounts. (Non-business versions of Microsoft 365 already include Copilot for free.)
Also free will be the ability to use Copilot to create agents to automate tasks and workflows, though advanced capabilities are only available at an extra cost.
Why is Microsoft doing it? It recognizes that companies can no longer charge for basic AI tools — there’s too much competition. Google Workspace, for example, already includes the use of its genAI tool, Gemini.
Intellectual property theft and AI training
GenAI tools like Copilot and ChatGPT require vast amounts of text to train them. OpenAI, Microsoft and other AI companies grab tremendous amounts of text from copyrighted material in newspapers, magazines, books, and other sources — and don’t pay the copyright owners for it.
Microsoft and other AI companies claim that doing so follows “fair use” guidelines for using copyrighted materials. Owners of the copyrighted materials call it theft. (I’m among the latter, because at least two dozen of my books have been used to train AI without my permission, and without payment. I’ll have more to say about that in a future column.)
There have already been many lawsuits about this issue. The most important one was brought in 2023 by The New York Times, nine other newspapers, and the Center for Investigative Research against Microsoft and OpenAI. Don’t expect that lawsuit to be settled in 2026. It might not even be heard in 2027.
And even if it is heard, expect appeals all the way to the US Supreme Court. The result? Microsoft, OpenAI, and other AI companies will be able to continue to take whatever content they want with impunity this year.