Context: Competition authorities in the Western hemisphere are increasingly looking at AI-related partnerships (May 21, 2024 ai fray article) as well as questions of unilateral conduct.
What’s new: Today the four leading Western competition enforcers (U.S. Department of Justice, U.S. Federal Trade Commission, European Commission and UK Competition & Markets Authority) issued a joint statement on Generative AI.
Direct impact: The agencies clarify upfront that they will continue to act and decide independently, and obviously operate under different legal frameworks, but they will compares notes in light of the inherently cross-border nature of Generative AI. The statement is presumably also meant to send out a strong signal that there will be more GenAI-related antitrust investigations.
Wider ramifications: In recent years, U.S. and UK competition “watchdogs” have seen eye-to-eye on digital industry matters to a greater extent than with their colleagues in Brussels, whom they regarded as soft. It is, therefore, even more interesting to see that a consensus was reached between “progressives” (U.S., UK) and “conservatives” (Brussels), albeit at a high level of abstraction.
There is an inherent difference between the remits of the European Commission’s Directorate-General for Competition (DG COMP) and the FTC, DOJ and CMA. The EC does not have a general consumer protection function. That is why the last paragraph of the statement (“Consumer risks associated with AI”) excludes the EC.
But before the document talks about “risks” of any kind, it recognizes that “[a]t their best, these technologies [AI-related ones such as foundation models] could materially benefit our citizens, boost innovation and drive economic growth.”
The statement identifies three main categories of risks to fair and effective competition:
- concentrated control of key inputs such as specialized chips, cloud computing power, large-scale data, and specialist technical expertise;
- entrenching or extending market power in AI-related markets by leveraging market power at different levels related to the AI technology stack; and
- arrangements involving key players (partnerships, financial investments, and “other connections between firms related to the development of generate AI”), though the statement does recognize that such arrangements don’t necessarily harm competition in each case.
There is also a section on “other competition risks associated with AI” such as price-fixing or collusion, or the sharing of competitively sensitive information between competitors.
What the competition enforcers want to see is fair dealing, interoperability, and choice. There is widespread consensus about those objectives. A few months ago, Microsoft stated its AI Access Principles “to promote innovation and competition in the new AI economy” (February 26, 2024 Microsoft corporate blog post).
Today’s joint statement by the four agencies shows that Generative AI is now at (or at least near) the top of their respective priority lists. It increases the likelihood that there will be parallel inquiries into some AI competition issues in all three jurisdictions. To a certain extent, those agencies are allowed to exchange information, and there is no question that they will coordinate to the extent they can (and to which it makes sense, given that they have to apply and respect different laws). They will try to give each other political air cover, enabling agency A to defend its actions by pointing to what agencies B and C are doing.
At the end of the day, however, the enormous potential of Generative AI does not move the legal goal posts. There are longstanding rules, and there is plenty of case law, as to what constitutes a merger or an abuse of a dominant market position. Competition authorities don’t make or change the law (much less now in the U.S. after the Supreme Court’s Loper Bright ruling that did away with the Chevron doctrine): it’s their job to faithfully apply it.
This is not meant to diminish the significance of the statement, but three of the four signatories are not very likely ot be in office anymore in a year from now. EU competition commissioner Margrethe Vestager is not going to form part of the next Commission. In the U.S., there are elections in November. Should Donald Trump win, then there will definitely be a new Assistant Attorney General at the top of the Antitrust Division of the DOJ and a new FTC chair. If the Democratic Party remains in power, it would not be surprising if the current Antitrust AAG, Jonathan Kanter, returned to private practice after one term. It’s possible that several of the signatories wanted to get this joint statement out now as part of their legacy.