Joint Statement on Automated Systems: A Reminder of the Potential Harms of AI

This week, the FTC and DOJ along with the Consumer Financial Protection Bureau, and the Equal Employment Opportunity Commission released a joint statement pledging the agencies’ commitment to enforce their respective laws and regulations and to uphold their commitment to the core principles of fairness, equity, and justice in relation to emerging automated systems. As artificial intelligence becomes increasingly common in our daily lives, the joint statement reiterates the agencies’ “resolve to monitor the development and use of automated systems and promote responsible innovation” and to “protect individuals’ rights regardless of whether legal violations occur through traditional means or advanced technologies.”

The agencies have previously expressed concerns about potential harmful uses of artificial intelligence and have demonstrated their focus on enforcing any unlawful conduct effectuated through automated systems. For example, the FTC previously issued a report to Congress outlining concerns that artificial intelligence tools can be inaccurate, biased and discriminatory by design. The FTC has also issued guidance that companies should not exaggerate the capabilities of artificial intelligence and should be aware of the risks that an artificial intelligence product may have on the market.

While artificial intelligence offers the potential to greatly advance and modernize many applications, the use of artificial intelligence also has the potential to perpetuate unlawful biases, automate unlawful discrimination, and produce other harmful outcomes. In this regard, the joint statement highlights the potential discrimination inherent in the use of automated systems, including through datasets that incorporate historical biases and the lack of transparency in many automated systems that can make it difficult for developers and consumers to know whether the system is fair. With respect to unfair competition practices in particular, the antitrust agencies have expressed concerns that the use of pricing software and adoption of pricing algorithms may facilitate tacit collusion. Indeed, in its first online marketplace prosecution, the DOJ brought price-fixing claims against a former e-commerce executive in connection with an alleged agreement of sellers on Amazon Marketplace to implement a price algorithm that would maintain a supracompetitive price for wall posters.

The antitrust laws have been criticized as being slow to adapt to emerging technologies, including artificial intelligence, and numerous antitrust scholars have proposed modifications to current antitrust laws and policy, including addressing the potential for tacit collusion through pricing algorithms. In the meantime, however, the basic principles of antitrust law apply equally to the conduct of individuals and entities perpetuated through automated systems, and the agencies have demonstrated a commitment to apply the existing laws to such conduct. With the agencies’ continued commitment to enforcing their respective laws for conduct involving artificial intelligence, we can expect similar cases going forward and an expansion of the type of conduct that the agencies pursue as technology advances. We can also expect that the agencies’ focus on potential harmful uses of artificial intelligence will cross over with the agencies’ other enforcement priorities, including, for example, unfair competition practices in the labor market.

Key Takeaways

While the agencies and courts grapple with balancing the benefits and harms in this evolving area, the joint statement is yet another reminder that the antitrust laws apply with equal force to automated technology and of the agencies’ commitment to pursue continued enforcement to any unlawful conduct leveraging such technology. Companies should therefore be mindful of how automated systems are designed and implemented to ensure that automated systems are set up in a fair and competitive manner, and should consult with antitrust counsel throughout the process.