Washington
CNN
—
The Federal Trade Commission is investigating Microsoft's recent deal with artificial intelligence startup Inflexion, according to people familiar with the matter, as U.S. antitrust regulators step up scrutiny of the hot-button AI industry.
In March, Microsoft announced it had hired Inflection's co-founder and several staff members to head its Copilot program, and said Inflection would host its AI models on Microsoft's cloud platform. As part of the deal, Microsoft reportedly paid Inflection $650 million. In a statement at the time, Microsoft described the move as simply a hiring decision, not an acquisition.
The FTC's investigation into Microsoft is about whether its investment in Inflexion constituted an acquisition that the company failed to report to the government, one of the people said.
The investigation comes as the FTC and Justice Department antitrust authorities are nearing a final agreement this week on how they will jointly oversee AI giants including Microsoft, Google, Nvidia and OpenAI, two people familiar with the matter told CNN.
The deal, which could be finalized within days, would name the Justice Department lead investigator for Nvidia and put the FTC in charge of investigating Microsoft and OpenAI, the people said. The Justice Department will likely continue to oversee Google, one of the people suggested. The investigation will focus on whether the companies used their dominance in AI to stifle competition through fraudulent or illegal conduct.
Microsoft declined to comment on the Justice Department and FTC settlement but defended its partnership with Inflexion in a statement.
“Our agreement with Inflection provides us the opportunity to hire talent at Inflection AI and build a team that can accelerate Microsoft Copilot, while allowing Inflection to continue pursuing its independent business and ambitions as an AI studio,” a Microsoft spokesperson said, adding that the company is “confident” it is in compliance with its reporting obligations.
Inflection and Google did not immediately respond to requests for comment. Nvidia and OpenAI declined to comment. The DOJ and FTC also declined to comment. The FTC-DOJ settlement was first reported by The New York Times, and the FTC investigation into Microsoft was first reported by The Wall Street Journal.
FTC Chairman Lina Khan has warned in op-eds and congressional testimony that AI, if left unchecked, could “accelerate” fraud and deception. The agency has issued multiple reminders and warnings that companies could be held liable if they make misleading claims about their AI tools or secretly use consumer data to train AI models. The FTC is currently investigating Reddit's AI content licensing practices and is separately investigating OpenAI for possible violations of consumer protection laws.
The U.S. agency separation of labor opens the door to more thorough scrutiny of a sector that has energized investors, enticed consumers, and raised alarms among critics who say AI is in urgent need of regulation to prevent widespread job losses, discrimination, and fraud. Specifically, it gives the FTC and DOJ the power to investigate whether big tech and AI companies are engaging in anti-competitive behavior.
It also highlights the growing trend for enforcement agencies to apply existing laws to the industry as the outlook for new laws regulating AI in the U.S. grows dim. The U.S. is widely seen as a laggard in regulating AI, and other countries, such as the European Union, have leapfrogged it with tougher rules on the use of the technology in high-risk situations. For example, EU AI law outlaws AI-powered social scoring systems and any biometric-based tools used to infer a person's race, political leanings, or sexual orientation. It also bans the use of AI to interpret people's emotions in schools and the workplace, as well as some automated profiling aimed at predicting the likelihood of them committing future crimes.
For years, technology critics and regulators have worried that big tech companies are dominating entire sectors of the economy, leading to several high-profile antitrust lawsuits from the U.S. government targeting Amazon, Apple, Google, Meta and Microsoft.
Some worry that tech companies will abuse their powerful role in business and society to extend their dominance into the burgeoning field of generative AI, which exploded with OpenAI's release of ChatGPT in 2022.
Nvidia's soaring stock price serves as a barometer of the AI craze, underscoring the company's position as a major supplier of computing chips needed to train advanced AI models. On Wednesday, Nvidia became the second-largest publicly traded company in the U.S., with a market capitalization of more than $3 trillion as of that day's closing price, surpassing Apple.
“AI relies on huge amounts of data and computing power, which could give huge advantages to companies that already have an advantage,” Jonathan Cantor, the Justice Department's antitrust director, said in a speech at Stanford University last week, adding that Americans' reliance on just a handful of major tech companies could enable those companies to “dominate these new markets.”
Critics say one way tech giants exert anticompetitive influence in the AI space is through exclusive partnerships with AI startups. These deals can “lock in” AI developers as customers of large-scale cloud computing services and give the tech giants a significant stake and influence over AI development. These types of agreements, including Microsoft's relationship with OpenAI, are the subject of an ongoing FTC investigation announced in January.
The Department of Justice has also become more vocal on AI issues: In 2022, its antitrust division hired Stanford University professor and AI expert Susan Athey as its chief economist.
When considering competition in the AI industry, antitrust enforcement agencies should look to past anticompetitive behavior by tech giants, Acey told CNN at a recent Bloomberg News event in Washington.
This could include gatekeeping or bottleneck tactics, making it difficult for consumers or customers to switch providers, or becoming the largest purchaser of critical supplies, like AI chips, for example, and denying needed supplies to competitors.
“We should look back at parallel examples in history and see where the sources of market power have been and how people have maintained it, and those are tactics we might be concerned about going forward,” Athey said.
This story has been updated to reflect additional background and developments.
