How OpenAI is boosting scrutiny of Microsoft’s market power  

Microsoft’s burgeoning relationship with OpenAI is piling scrutiny on the tech giant’s market power and ways it is building and yielding that power in the lucrative artificial intelligence (AI) space.

Microsoft successfully skirted some of the spotlight on the market power of other technology giants, such as Apple and Google, over the past few years. As artificial intelligence (AI) becomes a target for global regulators and lawmakers, Microsoft’s growing partnership with the leading AI firm and creator of ChatGPT is bringing renewed attention.

The Federal Trade Commission (FTC) is examining the nature of Microsoft’s investment in OpenAI, Bloomberg News reported. The U.K. Competition Market Authority (CMA) also launched an initial review last week to see if Microsoft’s growing relationship with OpenAI “resulted in a relevant merger situation,” and if that change would lessen competition.

Microsoft invested billions of dollars in OpenAI and has incorporated ChatGPT into Microsoft services. The lines between the two companies, however, have been blurred in the aftermath of the tumultuous firing and rehiring of OpenAI co-founder and CEO Sam Altman.

“The relationship between Microsoft and OpenAI existed prior to that. But that debate, that very public dispute, showed really the extent of Microsoft’s control over OpenAI,” said Lee Hepner, legal counsel at the American Economic Liberties Project, a nonprofit that supports aggressive antitrust enforcement.

The OpenAI board announced Altman was ousted as CEO on Nov. 17, a Friday. By the following Monday, Microsoft announced it would be hiring Altman to lead an AI research team.

Altman returned to the AI startup just a few days later — with the blessing of Microsoft CEO Satya Nadella — after hundreds of OpenAI employees threatened to quit if he was not reinstated.

One week later, OpenAI announced Microsoft would have a nonvoting position on its board of directors.

“The question on everyone’s minds here is, ‘Who absolutely controls OpenAI?’” Hepner said.  

“On the one hand, you have this perception of a nonprofit board that is controlling OpenAI and making critical decisions about the future of development of AI, generally. On the other hand … the public tussle over the disposition of Sam Altman and his role at OpenAI, I think, made very clear that actually Microsoft controls OpenAI,” Hepner added.  

Microsoft has pushed back on accusations that it controls OpenAI.

After U.K. regulators asked for comment on the Microsoft-OpenAI partnership — the first step before the potential launch of a formal investigation — Microsoft President Brad Smith said it has “fostered more AI innovation and competition, while preserving independence for both companies.” 

Smith said the “only thing that has changed is that Microsoft will now have a non-voting observer on OpenAI’s board.” He also said the situation is “very different from an acquisition such as Google’s purchase of DeepMind in the UK,” seemingly taking a shot at rival Google over its 2014 acquisition of an AI company.

Sam Weinstein, a former Justice Department antitrust attorney who is now a professor at the Cardozo School of Law, said the change in having a nonvoting board observer “potentially raises more red flags” for antitrust enforcers.

“Instead of what’s been characterized by Microsoft as this kind of economic interest where they have no actual control … now you have a board observer. Of course, they don’t have a vote, but they’re in the room. So do they exert control there? We won’t know — we’re outside the room — but it’s a question,” he said.

The situation also draws more questions about how information is flowing between the companies, he said.

“The board observer kind of raises the stakes a little bit and catches the eye of the enforcers,” Weinstein added, acknowledging the situation falls in a legal “gray area.”

The Clayton Act prohibits a board member from serving on the board of management of a competing corporation, but a nonvoting observer from Microsoft may not violate that law, he said.  

“One way you could look at this is Microsoft’s exerting some kind of control over OpenAI, it’s kind of hard to say how much, in a way that at least seems like it doesn’t implicate the antitrust laws,” Weinstein said.

“They’re sort of dancing around the scope of what the enforcement agencies can do by having this unusual arrangement,” he added.  

Microsoft has also sought to tamp down concerns about its influence on OpenAI.

“While details of our agreement remain confidential, it is important to note that Microsoft does not own any portion of OpenAI and is simply entitled to share of profit distributions,” Microsoft chief communication officer Frank Shaw said in a statement following the CMA’s invitation for comment.

The FTC declined to comment on whether it was probing the Microsoft-OpenAI relationship, but a source familiar with the situation told The Hill that the agency has not reached out to Microsoft about the matter.

FTC Chair Lina Khan, a vocal critic of tech giants’ market power before assuming her role at the agency, has focused the commission on the potential competition concerns raised by AI.

In a June memo, the agency pledged to use its “full range of tools to identify and address unfair methods of competition.”

The FTC voted unanimously last month to authorize the use of compulsory process, such as subpoenas, in nonpublic investigations involving AI-related products and services. The FTC is currently controlled by three Democratic commissioners after the departure of two Republican colleagues earlier this year and last.

The agency said the update “will streamline” the agency staff’s ability to issue civil investigative demands.

“There’s an obvious effort at the Federal Trade Commission to move quickly here, and I think that that’s what is most important. The outcome will be a result of the investigation, but it’s critical that the Federal Trade Commission move quickly before this runaway train gets too far down the track,” Hepner said.

The American Economic Liberties Project and other tech advocacy groups, including Demand Progress and the Center for Digital Democracy, sent a letter Wednesday to Khan and Jonathan Kanter, assistant attorney general and head of Justice Department antitrust division, urging their offices investigate “Big Tech’s concentration in the AI space.”

The letter specifically calls out changes between Microsoft and OpenAI’s relationship, including the restructuring that secured Microsoft a nonvoting board position. The groups said the change “indicates a greater level of access and control for Microsoft over the management and future trajectory of OpenAI,” according to a copy of the letter.

The rise of generative AI comes heels of the federal government’s ramped up scrutiny of the market power of tech giants — an effort that spawned a blockbuster 2020 House report and lawsuits filed by antitrust enforcers with both the Trump and Biden administrations.

While Microsoft avoided much of the recent push to rein in Big Tech, the company is no stranger to antitrust scrutiny.

The federal government sued Microsoft in 1998, alleging the company was violating antitrust law through restrictions placed on its software. Microsoft and the Justice Department eventually reached a settlement after the company successfully appealed a district court ruling that would have broken up the firm.

More recently, the FTC challenged Microsoft’s $69-billion acquisition of the gaming company Activision Blizzard, but the deal went through after the FTC lost its bid to block it.

The FTC is still fighting the merger, through an appeal, even after the deal closed. A decision on the appeal has not yet been reached.

The 2020 House antitrust report — and hearings with the CEOs of Apple, Amazon, Google and Facebook that followed — focused largely on how the tech giants amassed their power and if they stifled innovation by acquiring nascent companies as the modern-day internet and social media landscape took shape.

Microsoft may have avoided the spotlight of that push, but Weinstein said the industry is at another inflection point as key AI companies rise up and pose a potential threat — or valuable input — to tech giants.

Microsoft and OpenAI are now in the middle of that debate.  

Regulators have to face the question of whether companies like Microsoft and Google should be able to control their affiliated AI companies or if they should compete separately and possibly “overthrow the current set of dominant” firms, Weinstein said.  

“If you want the latter, and think ‘OK, these companies are the competitive threat we’ve been waiting for, to the Facebook’s, Apple’s, Amazon’s, etc., you probably want them outside the control of those monopolists, as pure competition matter. And that doesn’t seem to be what’s happening,” Weinstein said.