FTC warns AI industry on discrimination

Rep. Ted Lieu introduced a House bill Monday that would crack down on deepfake distribution, non-consensual AI imagery, and retaliation against AI whistleblowers. (cnbc.com) CNBC reported the measure first on April 27, 2026. Lieu said the bill is “a step forward” and said he wants Congress to pass something “this term right now.” (cnbc.com) The bill would increase penalties for distributing deepfake and non-consensual images, protect workers who report AI safety risks or legal violations, and require U.S. participation in international AI standards bodies. (cnbc.com) Lieu’s proposal also would create a prize competition for artificial-intelligence research and development. Rep. Jay Obernolte, the California Republican who co-led the House AI task force with Lieu, supports the bill while drafting a separate package of his own. (cnbc.com) A deepfake is synthetic audio, video, or imagery generated to look real. Lawmakers have focused first on non-consensual sexual images and impersonation because those harms are easier to define than broader questions about how to regulate general-purpose AI systems. (cnbc.com) That narrower approach leaves out two of the hardest fights in Washington: whether federal law should override state AI rules, and whether Congress should impose testing requirements for AI used in schools, infrastructure, or other sensitive settings. (cnbc.com) The Federal Trade Commission has been making a parallel argument for years: companies do not get a pass because a product uses AI. In a 2023 joint statement with the Justice Department, Equal Employment Opportunity Commission, and Consumer Financial Protection Bureau, the agency said existing laws already apply to discrimination and bias in automated systems. (ftc.gov 1) (ftc.gov 2) Former FTC Chair Lina Khan said at the time that “there is no AI exemption to the laws on the books.” In a separate statement, she said the FTC’s authority over unfair or deceptive practices “squarely applies” to AI systems that harm or mislead people. (ftc.gov 1) (ftc.gov 2) The agency has paired that warning with enforcement. In September 2024, the FTC announced “Operation AI Comply,” a sweep of five cases involving deceptive AI claims, fake reviews, and other allegedly unfair or misleading conduct. (ftc.gov) It kept going in 2025 and 2026. The FTC sued Air AI in August 2025 over claims that customers could earn money with its calling software, then said on March 24, 2026 that the company and its owners would be banned from marketing business opportunities under a settlement. (ftc.gov 1) (ftc.gov 2) The result is a two-track policy push on AI in Washington on April 27, 2026: Congress is trying to write its first narrower rules around deepfakes and whistleblowers, while the FTC is signaling it can already police bias, deception, and other harms with laws it has now. (cnbc.com) (ftc.gov)