Tech antitrust: Microsoft under FTC scrutiny while Google, Nvidia and AI labs face cases

Microsoft spent years watching its peers get hauled into court while it stayed out of the antitrust crosshairs. That grace period may be closing. The Verge reports that the Federal Trade Commission has sent civil investigative demands, which work like subpoenas, to at least half a dozen companies that compete with Microsoft. The agency's questions point at Azure cloud services and at Microsoft's position in AI.

The investigation began under the Biden administration in 2024 and continued under President Donald Trump, The Verge reported on 1 June 2026. Nothing guarantees it ends in a lawsuit. FTC staff will decide whether to recommend a complaint, and the agency's two commissioners vote on it.

What the FTC is asking about

The Verge obtained previously unreported details from an industry source who reviewed the letters and was granted anonymity to discuss nonpublic information. The documents generally run more than 15 pages and more than 15 questions, often with extensive sub-parts. The questions cover Microsoft's business agreements, licensing arrangements and product interoperability. They ask competitors about organizational charts, marketing plans and product roadmaps. Some probe how hard it is to break into markets where Microsoft operates, asking about the costs and barriers to entry. Others ask companies to detail plans for bundling features and products, plus pricing, discounting and profitability. According to the source, the CIDs say the goal is to determine whether Microsoft used unfair methods of competition in cloud, software products and related services in violation of the FTC Act. The FTC did not respond to a request for comment.

Microsoft spokesperson Alex Haurek says the company is "cooperating fully with the FTC and believe our practices promote competition while delivering the innovative products our customers expect."

Customers have complained, often anonymously for fear of retaliation, that Microsoft's 2019 licensing changes made it significantly more costly to run Windows software outside Azure. In 2023, Google told the FTC in a broad cloud market inquiry that Microsoft used dominance in other areas to give its own cloud products an unearned advantage and lock in consumers. Microsoft disputes that framing. Haurek pointed to Google as "a clear example of that dynamic, growing 63% year over year and competing head-to-head with other major providers including AWS, the largest cloud provider."

George Washington University law professor and former FTC chair William Kovacic told The Verge this kind of information gathering tends to happen early in an investigation. He said the questions suggest the agency is trying to work out which market players to watch and how important tactics like bundling are. The AI references, he said, could indicate the FTC sees AI as part of a possible illegal tying arrangement or as an important competitive advantage.

The US is not alone. The Verge notes the European Commission, the UK Competition and Markets Authority and Japan Fair Trade Commission have also been probing the issue. A 2024 ProPublica investigation, cited by The Verge, found Microsoft used similar leveraging to gain a foothold across US government agencies after installing free cybersecurity upgrades tied to products that ran on Azure. Government contracts and cybersecurity were not among the CID topics disclosed to The Verge. Haurek said Microsoft responded to the government's security requests at the time and that agencies were "free to engage with other vendors."

Google's remedy phase

The Microsoft probe is happening while older cases move into their punishment stages. In Google's search antitrust trial, the liability phase is already over. Ars Technica reported on 22 April 2025 that the court found Google violated the Sherman Antitrust Act by willfully acquiring and maintaining monopoly power in search services and general search advertising.

The remedy phase, overseen by US District Judge Amit Mehta, is about what to do next. The Department of Justice is seeking to force Google to divest Chrome, unbundle Android and end the search placement deals that pay Apple and Mozilla billions of dollars a year to make Google the default search engine. Ars Technica reported that the DOJ sees this as the most consequential US antitrust case since the Microsoft trial of the 1990s, and that it plans to call witnesses from Yahoo, DuckDuckGo and Microsoft to describe how their search businesses were stymied.

Google's counter-argument is that the government's position is too extreme. The company has proposed less exclusivity in search contracts and more freedom for Android device makers to choose app preloads, plus additional government oversight. It will also argue that Chrome's data sharing demands threaten user privacy and that Chrome, built on open source Chromium, has been run as a benevolent steward of the open web.

Ars Technica noted that Mozilla and Apple, both paid billions per year under the default deals, will testify that Google's services are the best available. The DOJ is likely to counter that this revenue makes their testimony less reliable. Mozilla's near-total dependence on Google money is not disputed, and Google has argued in the past that Mozilla would fold without it.

Nvidia's China problem

Outside the US, the most concrete ruling so far landed on Nvidia. Ars Technica reported on 15 September 2025 that China's State Administration for Market Regulation made a preliminary finding that Nvidia violated the country's antitrust law by failing to fully comply with provisions set when it acquired Mellanox Technologies.

Beijing conditionally approved the $6.9 billion Mellanox acquisition in 2020. Two people with knowledge of the matter told Ars Technica that SAMR reached its conclusion weeks before the announcement and released the statement to give China greater leverage in trade talks with the US in Madrid. The regulator opened the anti-monopoly investigation in December, a week after Washington unveiled tougher export controls on advanced high-bandwidth memory chips and chipmaking equipment.

The potential penalties are large. Fines could run between 1 percent and 10 percent of Nvidia's previous year's sales, and regulators can force changes to business practices they consider illegal. Nvidia did not immediately respond to a request for comment, and SAMR did not immediately respond to a request for further comment.

Nvidia has been squeezed between Washington and Beijing for years. US export controls forced it to sell watered-down versions of its AI graphics processors in China, creating a black market for more advanced chips. The Trump administration blocked sales of the H20 chip, designed for China under export rules, then allowed sales to resume in exchange for 15 percent of revenues going to the US government. Chinese regulators have warned domestic companies not to buy the H20, adding uncertainty to Nvidia's China business.

The AI labs in the dock

Meanwhile, the companies building frontier AI models are facing a different kind of claim. Tom's Hardware reported on 21 September 2026 that four plaintiffs who subscribed to ChatGPT, Claude, Grok or Gemini filed a proposed class action alleging that the model developers violated antitrust laws by agreeing to slow AI development.

According to the Associated Press, cited by Tom's Hardware, the suit argues the agreement would reduce the value consumers get from paid AI subscriptions. The plaintiffs say the coordination started in July 2026, after leading AI labs signed a statement admitting there is intense competitive pressure not to unilaterally slow development. The plaintiffs accept the safety rationale but call Anthropic founder Dario Amodei's cooperation proposal a shortcut that substitutes collective restraint for individual accountability.

Attorney Nick Rowley, lead counsel for the plaintiffs, says: "AI will quickly spin out of human control and could kill us all if we allow AI safety and protocol ... to be controlled by private self-serving agreements between the world's most powerful 'for profit' technology companies."

Amodei's essay acknowledged the antitrust risk and suggested he hoped the government would grant an exception. OpenAI's Sam Altman responded on X, saying: "We welcome a federal framework that sets consistent safety requirements for frontier AI. But we do not believe we need to wait for an antitrust exemption or legislation to begin the work of providing this confidence." The Trump administration rejected the idea. The president said: "AI taking over the World, destroying Humanity, and all other things bad, is a HOAX."

Chinese state media also criticized the proposal. China Daily called it a club whose membership rules have been drafted before the guest list is announced, and argued that a global AI safety framework that excludes China is not quite global. Amodei's essay had explicitly mentioned slowing China's progress and widening the US lead over Beijing.

Why the timing matters

The four cases sit at different stages and carry different risks. Microsoft is at the information-gathering stage, the earliest point, where the commission could quietly drop the matter. Google is past liability and arguing over remedies that could change how search distribution works. Nvidia faces a preliminary foreign finding that could translate into fines and forced practice changes. The AI class action, filed in September 2026, has yet to be tested in court.

What connects them is the widening definition of antitrust harm. Cloud licensing terms, default search deals, chip export workarounds and pledges to slow model development are now all being examined as possible restraints on competition. Whether any of it produces a breakup or a fine is still open. The questions, for now, are doing more work than the answers.