No Robot Bosses Act

Key Points

• The bill would ban employers from relying solely on AI or automated systems to make hiring, firing, or disciplinary decisions. Every major workplace decision involving an algorithm must be reviewed and approved by a qualified human being.
• Employers with 11 or more workers would be required to disclose when they use AI in the workplace, explain what data the system collects, and describe how the system reaches its conclusions — all in plain language, at no cost to the worker.
• Workers would gain the right to dispute any AI-generated decision and appeal it to a human — and employees managed by automated systems could opt out entirely and request a human supervisor instead.
• Before deploying these systems, employers must test them for discriminatory bias based on race, sex, age, disability, and other protected characteristics. These bias audits must be repeated annually and made public.
• A new Technology and Worker Protection Division would be created within the Department of Labor to enforce these rules. Workers who are harmed could sue for $5,000 to $40,000 per violation, and forced arbitration clauses would be unenforceable for these disputes.
• State attorneys general and state privacy regulators could also bring enforcement actions on behalf of their residents, with statutory damages of up to $50,000 per violation — or $100,000 for repeat offenders who cause serious economic harm like firing someone.