Artificial intelligence in the workplace

(a) Establishment Using amounts available for such purpose from the Artificial Intelligence Horizon Fund, the Secretary of Labor, in consultation with the Equal Employment Opportunity Commission, shall establish an initiative to develop, disseminate, and periodically update best practices and guidance for the safe, responsible, and effective integration of artificial intelligence technologies in the workplace. (b) Convening In carrying out the initiative established under subsection (a), the Secretary shall convene representatives of— (1) employers; (2) human resource professionals and organizations representing human resource professionals; (3) labor organizations; (4) civil rights organizations; (5) institutions of higher education; (6) cybersecurity professions; (7) legal professionals responsible for advising employers on workforce and employment matters; and (8) such other public and private stakeholders as the Secretary determines appropriate. (c) Duties In carrying out the initiative established under subsection (a), the Secretary, in consultation with the stakeholders described in subsection (b), shall— (1) develop voluntary guidelines for employers regarding the responsible adoption and use of artificial intelligence technologies in the workplace; (2) develop model policies and implementation resources for employers; (3) identify practices to mitigate risks relating to worker privacy, bias, discrimination, transparency, and accountability in workplace uses of artificial intelligence, including practices related to— (A) worker and labor organization participation in the design, procurement, testing, and evaluation of artificial intelligence; (B) advance notice to, and consultation with, workers and labor organizations before the adoption of artificial intelligence technologies that materially affect working conditions; (C) algorithmic management; (D) workplace surveillance, including limits on the collection and use of worker data; (E) automated employment decision tools; (F) disclosure to workers related to the use of artificial intelligence; (G) human review of consequential employment decisions, including decisions relating to hiring, scheduling, discipline, compensation, and termination; (H) the ability of a worker to exercise professional judgment to override an automated determination, and protection from retaliation for doing so; (I) the ability of a worker to meaningfully challenge or appeal employer decisions made using artificial intelligence, including the provision to the worker of an understandable explanation of the basis for an adverse decision and a process to request human reconsideration; (J) protection from retaliation for a worker who reports a harm or misuse relating to workplace artificial intelligence; (K) the identification and reporting by employers of workforce reductions substantially related to the adoption of artificial intelligence; (L) independent testing for bias and discriminatory outcomes of artificial intelligence systems, before and during deployment; and (M) the use of artificial intelligence to augment the capabilities of workers and improve job quality and workplace health and safety; (4) periodically review and update the guidelines to reflect technological advances and emerging best practices; (5) make all guidance, model policies, and related resources publicly available; and (6) not later than 2 years after the date of enactment of this Act, and periodically thereafter, submit to the Committee on Health, Education, Labor, and Pensions of the Senate and the Committee on Education and Workforce of the House of Representatives a report identifying which of the practices described in paragraph (3) should be established as requirements, including any recommendation for legislation and an assessment of the extent to which voluntary adoption of such practices has occurred. (d) Federal agencies In carrying out the initiative established under subsection (a), the Secretary shall develop guidance under this section that is suitable for use by Federal agencies in adopting artificial intelligence technologies affecting the work of their employees or the delivery of public services. (e) Rule of construction Nothing in this section shall be construed to alter any right or obligation under Federal labor or employment law or to affect the terms of a collective bargaining agreement.