With great promise comes great scrutiny. As artificial intelligence (“AI”) has become part of industries’ and individuals’ daily repertoire, it has also come under focus by antitrust regulators. The DOJ, in its so-called “Project Gretzky,” is gearing up with data scientists and others to be a tech-savvy version of its former self. And it has made its position clear in recent filings: “Antitrust law does not become obsolete simply because conspirators find new ways to act in concert.” And this past summer, the FTC sent a 20-page CID to OpenAI, the maker of generative AI chatbot ChatGPT, apparently looking into potential unfair practices and “reputational harm” to consumers.
But as to competition and antitrust, the agencies’ focus appears to be on four primary areas:
- Control of Data. One of the central issues with AI is control of data, as AI development is incredibly resource-intensive, with large amounts of high-quality data necessary to train robust models. To this end, FTC Chair Khan warned that “a handful of powerful businesses control the necessary raw materials that start-ups and other companies rely on to develop and deploy A.I. tools.” Regulators will look closely at control over data, whether through contract or otherwise, and its impact on markets and competition.
- Cloud Computing. Large-scale computing power is essential—particularly for building pre-trained base models that form the foundation of all generative AI tools. On November 8, 2023, the G7 competition regulators recognized that “[s]ignificant computational resources such as cloud computing services and large-scale computing power” are critical inputs “for the development and deployment of large-scale AI.” The high cost of acquiring or renting compute resources may put this key input out of reach for new market entrants, and may create temptation for better-resourced incumbents to engage in self-preferential treatment. The FTC has also said it will focus on potential monopolies and consolidation in cloud services and computing.
- HR. The DOJ has warned that AI might be used to surveil workers and share information, facilitating blacklisting or boycotting of employees. The DOJ has more generally had its sights on information sharing that can allegedly be used to share employee salaries and depress their wages across an industry.
- Collusion. Both the FTC and DOJ have noted that AI tools could potentially be used for collusion through anticompetitive information sharing, including in lower concentrated markets. The agencies put out a white paper saying, “If competing firms each entered into separate agreements with a single firm [like a platform] to use a particular pricing algorithm,” with the understanding that others in the industry were doing the same, it could be the basis of an information sharing or hub and spokes conspiracy.
This, of course, is only the beginning. The continued development and deployment of AI tools is all but inevitable. The agencies themselves are looking into AI-powered tools to supercharge their investigative capabilities. As we press on, we also remain vigilant.