Federal Agencies Aim to Combat Algorithmic Bias in AI

Auditing has a role in preventing, investigating, monitoring and remedying algorithmic bias in artificial intelligence (AI), technologists and researchers from federal agency civil rights offices said during a meeting held Wednesday (Oct. 9).

The meeting was convened by the Justice Department’s Civil Rights Division to promote the coordination of AI and civil rights, the department said in a readout of the event that was released Friday (Oct. 11).

“Auditing is used to verify that algorithms generate accurate results, as opposed to reflecting historical bias against protected classes,” the readout said.

The meeting’s participants also discussed their recent developments in enforcement, policy initiatives, rulemaking, education and outreach related to safeguarding civil rights, according to the readout.

These include a Federal Trade Commission (FTC) report finding that social media and video streaming companies engaged in “vast surveillance” of their users, with insufficient privacy controls; an Equal Employment Opportunity Commission (EEOC) report that called for efforts to address barriers to equal opportunity in the tech sector; a Department of Labor (DOL) resource aimed at helping employers consider disability inclusion and accessibility in AI hiring technologies; and a Department of Education guide designed to remind developers of AI for education about their responsibilities to advance equity and protect civil rights, per the readout.

“All participants pledged to continue collaboration to protect the American public against any harm that might result from the increased use and reliance an AI, algorithms and other advanced technologies,” the readout said. “The agencies also agreed to partner on external stakeholder engagement around their collective efforts to advance equity and civil rights in AI.”

Under an executive order issued by President Joe Biden in October 2023, the Justice Department’s Civil Rights Division is tasked with coordinating federal agencies to combat unlawful discrimination and other harms that may result from the use of AI, according to the readout. This was the fourth such meeting called by the Civil Rights Division since that time.

The executive order also includes rules aimed at ensuring safe AI development, guarding against the threat of using the technology to engineer dangerous biological materials and promoting standards and best practices to detect AI-generated content and authenticate official content.