Four prominent horsemen of the Biden administration’s bureaucratic apocalypse—the Federal Trade Commission (FTC), U.S. Justice Department (DOJ) Civil Rights Division (DOJ), Consumer Financial Protection Bureau (CFPB), and the U.S. Equal Employment Opportunity Commission (EEOC)—came together April 25 to issue a joint statement pledging vigorous enforcement against illegal activity perpetrated through the use of artificial intelligence (AI) and automated systems.
AI is, of course, very much in the news these days. And when AI is used to violate the law, it obviously is fully subject to enforcement scrutiny. But why make a big splash announcement merely to state a truism?
One suspects there is more to the story. The language of the joint statement, together with the FTC’s accompanying press release, provide some hints. Those hints point to a campaign by the administration to effectuate de facto bureaucratic regulation of AI through overly expansive interpretations of existing law. The following discussion will focus on the FTC’s role in this initiative.
Discussion
The FTC’s brief press release embodies a broad view of AI-related wrongdoing. It notes that the four agencies “pledged today to uphold America’s commitment to the core principles of fairness, equality, and justice” as emerging automated systems, including AI, “become increasingly common in our daily lives – impacting civil rights, fair competition, consumer protection, and equal opportunity.” The release adds that the agencies have “resolved to vigorously enforce their collective authorities and to monitor the development and use of automated systems.”
The FTC’s references to ”fairness” and “fair competition” by implication allude to the fatally flawed November 2022 FTC Policy Statement on Unfair Methods of Competition (UMC). That policy statement has been roundly criticized (see the thoughtful critiques in the Truth on the Market symposium on the UMC statement) for rejecting the venerable consumer-welfare standard that had long guided FTC competition-enforcement policy, and replacing it with subjective notions of “unfair” conduct that could arbitrarily be invoked by the Commission to attack any conduct it found distasteful. (See then-Commissioner Christine Wilson’s dissenting statement.) Such an approach undermines the rule of law, ignores efficiencies, promotes uncertainty, and thereby harmfully interferes with welfare-promoting business conduct.
The specter of arbitrary FTC challenges to AI-related competitive practices that are misunderstood by the commission is obvious. Arbitrary legal attacks on AI practices on dubious subjective grounds could forestall a substantial amount of welfare-generating innovation in the AI space. This would reduce economic wealth creation and harm American technological progress in AI, in addition to weakening the U.S. efforts to prevent China from becoming dominant in this key realm (see here for a discussion of the U.S.-China AI rivalry).
The statement’s announcement that the agencies intend “to monitor the development and use of automated systems” is likewise troublesome. In the FTC’s case, it suggests a potential interest in deciding what forms of AI “development and use” are appropriate. Although rulemaking is not mentioned, the threat of litigation being brought by one or more of the agencies against certain disfavored AI implementations is real.
In particular, the threat of FTC UMC investigations and prosecutions could shape the nature of AI research by directing it away from innovations that the commission dislikes. This would be a form of “regulation by enforcement oversight” that could substantially slow progress in AI and thereby reduce economic growth.
The joint statement reinforces this problematic reading of the FTC’s press release. It stresses the FTC’s finding that:
AI tools can be inaccurate, biased, and discriminatory by design and incentivize relying on increasingly invasive forms of commercial surveillance.
The FTC, however, lacks a general statutory authority to combat “discrimination,” and its authority to attack forms of commercial surveillance likewise is highly dubious. The FTC’s proposed commercial surveillance and data security rulemaking, for example, flunks cost-benefit analysis and has other flaws that would prevent it from passing legal muster; see more here.
The notion that the FTC may challenge AI innovations it disfavors by bringing new questionable “discrimination” suits, and by concocting legally indefensible rule-based surveillance and data-security obligations, is a source of serious concern. As in the case of the UMC policy statement, the FTC would be taking novel actions beyond the scope of its congressionally granted authorities. Even if the courts eventually rejected such FTC initiatives, the costs reflected in foregone welfare-enhancing improvements in AI capabilities would be considerable.
The joint statement’s discussion of the CFPB, EEOC, and the DOJ Civil Rights Division less obviously supports the proposition that those agencies will be encouraged to act beyond their statutory mandates. It is notable, however, that various commentators have raised concerns about regulatory overreach by these three entities; with regard to the CFPB, for example, see here, here, and here.
Nevertheless, it is concerning that the administration would assign high priority to oversight of AI—an area of enormous technological and economic potential—to agencies that are concerned primarily with civil-rights issues and with consumer protection in the realm of financial services. The potential for regulatory mission creep that would harm American AI development and the dynamic competition it sparks is obvious.
Conclusion
The joint statement on AI and automated systems should be seen as a yellow (if not a red) warning flag that Biden administration efforts to micromanage AI development may be in the works. Particular attention should focus on the FTC, which has the potential to seriously undermine beneficial AI development through ill-conceived litigation and regulatory initiatives.
This is a serious matter. AI is of major consequence in the global political economy, particularly given China’s interest in the field. One can only hope that the FTC and the Biden administration will keep this sober reality in mind before they gin up new misguided forms of regulatory interference in the evolution of AI.