For Microsoft, 2025 could be a game-changing year

It’s looking like 2025 will be one of the most consequential years for Microsoft in a long time.  Since Satya Nadella became CEO in early 2014, the company has been on an upward trajectory, despite a few bumps along the way. Today it’s the most powerful AI company on the planet and the world’s third-most-valuable company, worth more than $3 trillion.

But in these volatile times, that can change quickly thanks to a still evolving AI market, the federal government targeting high tech, and the coming wild-card Trump presidency. What challenges will the company face in 2025, and how will it handle them? 

Here are my top five.

Microsoft and OpenAI will go from frenemies into enemies

Not so long ago, the Microsoft-OpenAI relationship was tech’s biggest bromance. Microsoft invested $13 billion in OpenAI, OpenAI’s influence and valuation skyrocketed, and Microsoft used the company’s generative AI (genAI) technology to vault to the top of the AI heap.

Last year, the bromance soured and the companies became frenemies; as OpenAI openly courted major Microsoft clients, Microsoft laid the groundwork for developing its own AI technology. Nadella disparaged OpenAI, saying, “If OpenAI disappeared tomorrow…, we have all the IP rights and all the capability. We have the people, we have the compute, we have the data, we have everything. We are below them, above them, around them.”

Don’t be surprised if there’s open warfare between the companies this year. OpenAI will be transforming itself from non-profit to a for-profit company, possibly changing the terms of its contract with Microsoft and allowing it to more easily pursue partnerships with other companies. In addition, the terms of the Microsoft-OpenAI deal says that when OpenAI’s ChatGPT achieves what’s called AGI and can reason on its own, Microsoft will lose its stake in the company. 

OpenAI CEO Sam Altman keeps hinting that’s coming sooner rather than later. Meanwhile,  Microsoft has been busy building AI technology that could replace OpenAI’s as the basis for Copilot and other AI products. 

The upshot? Expect open warfare between the two.

Microsoft will get hit with at least one US government antitrust suit

Decades ago, Microsoft was Big Tech’s bad boy, mowing down competitors with shady actions that drew the wrath of the federal government — an antitrust suit that dropped the company from the top tier of tech and led to a lost decade in which it became an also-ran.

After Nadella took the helm, Microsoft became Big Tech’s choirboy, largely avoiding any federal suits, while Amazon, Meta, Google, and Apple were hit with antitrust actions that threaten the core of their businesses. 

That will probably change in 2025. The Federal Trade Commission (FTC) has launched a  wide-ranging investigation into what it believes may be Microsoft’s anticompetitive practices. The agency is looking at the very heart of the company and its business practices — AI, cloud computing, its productivity suite, and Teams. The suit could also endanger the company’s billion-dollar contracts with the US government, because the FTC began looking at the company thanks to its poor security practices.

What’s unclear is whether the Trump administration will continue the investigation, and ultimately prosecute the company. My bet is it will. Top Trump advisor Elon Musk has become a Microsoft competitor with his AI company xAI, and he’ll most certainly use his high-level access to push for prosecution. He’s already suing Microsoft and OpenAI for trying to use their power to get a monopoly on AI.

Intellectual property battles will come to a head

Microsoft and other genAI companies face an even bigger problem than antitrust lawsuits — the lack of content on which to train genAI tools like Copilot and ChatGPT. Improving them requires massive amounts of intellectual property. So far, the companies have simply hoovered up anything they can find, largely without paying, claiming they can use the material under fair-use doctrine.

That’s led to plenty of lawsuits against Microsoft and other AI companies for intellectual property theft. In one of the biggest, The New York Times is seeking “billions of dollars in statutory and actual damages” because of what it calls the “unlawful copying and use of The Times’ uniquely valuable works.” 

Microsoft and other AI companies have begun making deals with publishers to pay for the content to train their AI models. In November, Microsoft inked a deal with the publisher HarperCollins in which it can use many of the company’s nonfiction books to train a new genAI product.

That deal might well be Microsoft’s first in a series of similar agreements with other companies. Expect more to follow this year.

Nadella will try to keep Trump at a distance

Since Donald J. Trump’s election, a number of Big Tech executives and companies have gone full-blown MAGA. In one of the more extreme makeovers, Meta’s Mark Zuckerberg has gone all in, donating $1 million to Trump’s inauguration, eliminating fact-checking on Meta platforms, backing away from policing hate speech, ending the company’s diversity efforts, killing transgender and nonbinary themes from its apps, and even removing tampons from its men’s bathrooms, which it had provided for nonbinary and transgender employees. 

He’s not the only one. Top executives from Google, Amazon, Apple, and others have also bowed to Trump. Microsoft had been the lone holdout until early January, when it donated $1 million to Trump’s inaugural. Aside from that, though, the company hasn’t curtailed diversity efforts or in any other way changed its culture to put it in line with Trump’s way of thinking. And Nadella hasn’t visited Trump in Mar-a-Lago, as have so many other tech execs.

The big question: Will Nadella continue keeping Trump at arm’s length, and not try to make the company culture more like Trump would like to see it? My guess is he’ll stay the course. But we’ll see.

We may get real revenue numbers for AI… or not

Finally comes perhaps the biggest issue of all for the company’s financial health: Can Microsoft sign up enough customers to make its genAI ambitions worth its while? We’re no longer in the early hype days when mere possibilities were more important than revenue. In 2025, the ROI rubber will meet the road of reality.

This is particularly important because the investments that need to be made in genAI are far greater in scale than in any technology before it. Microsoft can spend all the billions it wants for infrastructure, electricity, data centers, training, and development. But if businesses and people don’t find AI useful and open their wallets for it, that will mean nothing.

Microsoft claims it’s got plenty of customers, but it isn’t giving out details, such as how many people pay for Copilot on a monthly basis and how much revenue it gets from that. Instead, it publicizes potentially misleading statistics such as “More than 85% of the Fortune 500 are using Microsoft AI” and “Nearly 70% of the Fortune 500 are using Microsoft 365 Copilot.” Most likely those numbers are based on companies launching small pilot programs testing whether AI is useful. Pilots don’t bring in much revenue — only full deployments do.

We’ll know in 2025 if AI is starting to pay off when Microsoft touts real revenue numbers or the number of people actively paying for AI subscriptions. Until then, consider those numbers smoke and mirrors.