Earlier today, the EEOC, DOJ, CFPB, and FTC released a joint statement on “enforcement efforts to protect the public from artificial intelligence (AI) and automated systems.” The statement noted that while these tools offer the “promise of advance, their use also has the potential to perpetuate unlawful bias, automate unlawful discrimination, and produced other harmful outcomes.”
Specifically with respect to employment, the statement emphasized the EEOC and DOJ’s roles in enforcing federal anti-discrimination laws and pointed to the EEOC’s recent efforts on the subject, including enforcement activities on discrimination related to AI and automated systems and its issuance of a technical assistance document explaining how the Americans with Disabilities Act applies to the use of software, algorithms, and AI to make employment-related decisions about job applicants and employees.
Finally, the statement flagged a series of potential sources of discrimination that may violate federal law with respect to AI, including the unintended consequences of using unrepresentative or skewed datasets, the difficulty people may have identifying unfairness in an automated system given the typical “black box” nature of their internal workings, and the disconnect between developers’ assumptions about intended use or users of a system and the context in which those systems are ultimately used, by whom, and for what purpose.
Employers should take note of the increasing attention federal (and local Post | LinkedIn) authorities are paying to these systems and be mindful of the ways in which the use of AI or automated systems may pose risks to their business.