LegisTrack
Back to all bills
S 1792119th CongressIn Committee

AI Whistleblower Protection Act

Introduced: May 15, 2025
Labor & EmploymentTechnology & Innovation
Standard Summary
Comprehensive overview in 1-2 paragraphs

The AI Whistleblower Protection Act would make it unlawful for employers to retaliate against workers who report AI security vulnerabilities or AI-related violations. It covers both employees and independent contractors and applies to when the whistleblower provides information to regulators, the Attorney General, Congress, or supervisors who can address the issue. The bill creates a dual enforcement path: a complaint process with the Department of Labor and, if needed, a private civil action in federal court. Remedies include reinstatement, double back pay with interest, and compensatory damages (including litigation costs). Importantly, it bars any pre-dispute waivers or arbitration that would waive these rights. The act defines AI broadly—encompassing systems that learn or act with human-like capabilities and that use data-driven algorithms—while carving out typical consumer products embedded with AI (e.g., some word processors or navigation tools). It also sets thresholds for what counts as an “emerging AI technology” and defines what constitutes an AI security vulnerability or AI violation, including violations of federal law during AI development or deployment, or failure to respond to significant dangers AI may pose to safety, health, or national security. Overall, the bill aims to encourage reporting of AI risks by providing strong protections and a clear remedies framework.

Key Points

  • 1Broad protections for whistleblowers: Prohibits retaliation against covered individuals (employees and independent contractors) who report AI security vulnerabilities or AI violations or assist in related investigations or congressional inquiries. Protection extends to acting through regulators, law enforcement, Congress, or internal supervisors who can address misconduct.
  • 2Defined terms for AI scope: Establishes what counts as AI, artificial systems, and emerging AI technology, with particular attention to security vulnerabilities and violations tied to AI development, deployment, or use. Excludes standard, common commercial products embedded with AI from certain definitions to avoid overly broad coverage.
  • 3Enforcement and procedures: A whistleblower can pursue relief by filing with the Secretary of Labor, and if no final decision in 180 days (absent bad faith), may sue the employer in federal court. The case can involve a jury and follows procedures tied to a specific federal whistleblower law referenced in the bill.
  • 4Remedies for prevailing whistleblowers: Reinstatement with the same seniority, double back pay plus interest, compensatory damages (including litigation costs and attorney fees), and any other appropriate remedies as determined by the administering agency or court.
  • 5Nonwaivable rights: The bill prohibits waivers or alteration of these rights through contracts, collectively bargained agreements, or arbitration clauses, ensuring that whistleblowers retain access to the protections regardless of dispute-resolution provisions.

Impact Areas

Primary group/area affected: Workers in AI-related fields (employees and independent contractors) across industries that use or develop artificial intelligence, who might report vulnerabilities or violations without fear of retaliation.Secondary group/area affected: Employers and organizations that deploy or manage AI systems, including those who employ or contract with whistleblowers; regulatory and legal compliance frameworks may need updating to align with the new protections.Additional impacts: Could influence how AI security issues are reported and addressed, potentially increasing regulatory oversight and disclosures related to AI risks; may interact with existing whistleblower protections under other labor and safety laws; possible compliance costs for employers to ensure policies and training reflect these protections and to avoid inadvertent retaliation.
Generated by gpt-5-nano on Oct 7, 2025