Companies using artificial intelligence tools to hire, manage or monitor their employees could face new regulations in several US states led by Democrats if proposed legislation is enacted. Lawmakers in at least seven states have introduced more than a dozen bills this year aimed at setting rules around employers’ use of AI, from requiring more disclosures and transparency to mandating risk assessments to curbing worker surveillance.
US companies using artificial intelligence tools to manage their employees could face a range of new regulations in California, Connecticut and New York if proposed legislation is enacted. With more companies deploying AI in the workplace to vet potential hires, improve efficiency or cut costs, state legislators across the country are proposing a wide range of bills with hopes to address growing concerns about the potential harm to employees.
Lawmakers in at least seven US states have introduced more than a dozen bills this year aimed at setting rules around employers’ use of AI, from requiring more disclosures and transparency to mandating risk assessments to curbing worker surveillance.
For years, a top priority for states has been preventing discrimination in the workplace though the use of automated employment decision tools, with lawmakers pointing to several examples from corporate America, such as Amazon decommissioning its machine learning tool for recruiting in 2018 after it demonstrated a bias against women.
New York City passed the first rule in 2021 requiring independent audits of automated employment decision tools, and other states quickly followed.
Last year, Colorado passed the first law in the US that broadly regulates AI and includes provisions specifically protecting employees from algorithmic discrimination. Under the Colorado Artificial Intelligence Act, which takes effect next February, employers must notify employees or applicants about their use of AI. Illinois and Maryland have also enacted legislation that requires employers to obtain consent before using AI during certain parts of the hiring process.
This year at least 20 other states proposed comprehensive high-risk AI bills modeled after the Colorado AI Act that also include provisions aimed at preventing discrimination in the workplace, particularly during the hiring process. Many, however, have stalled after facing opposition from the tech industry in states such as Maryland and New Mexico. A Virginia bill that passed the legislature was vetoed by the governor (see here).
But more narrow proposed legislation related more specifically to the use of AI in the workplace is still pending, with lawmakers in California and New York among the most active.
A pair of New York bills (S 185 and A 3779) would expand New York City’s existing rules requiring independent audits of automated employment decision tools by setting up additional guardrails and guidelines for their use and adding new disclosure requirements for companies.
A bill from California Assemblymember Rebecca Bauer-Kahan (AB 1018) also places broad compliance obligations on both employers and AI vendors by requiring audits, data-retention policies and comprehensive risk assessments for AI-driven hiring tools.
Senator Jerry McNerney, a Democrat from Stockton, has introduced the No Robo Bosses Act (SB 7), which would prevent California employers from relying solely on automated decision-making systems (ADS) for hiring, promotions, discipline or termination. It would also ban AI systems from using workers’ personal data to predict their behavior. McNerney had helped set federal AI policy when he served in Congress from 2007 until 2023, where he co-founded the Artificial Intelligence Caucus and authored the AI in Government Act.
But the proposal could conflict with other state rules and regulations. SB 7 would impose requirements much like new rules also being proposed by the California Privacy Protection Agency related to automated decision-making technology. However, the CPPA rules only apply to for-profit businesses, and SB 7 would apply to all employers, including small businesses and nonprofits.
SB 7 could also overlap with anticipated new rules from California Civil Rights Council, which recently approved amendments to the state’s Fair Employment and Housing Act prohibiting the use of AI in certain hiring and employment practices conducted by businesses, nonprofits and governments (see here).
Other state lawmakers have introduced bills focused on specific workplace issues, such as AI surveillance tools or the collection of personal information.
Los Angles state Senator Lola Smallwood-Cuevas, a Democrat, has proposed SB 238, which would require all employers to notify the state annually of all AI surveillance tools used in the workplace, and what data they collect from workers.
In Massachusetts, state lawmakers in both the House and Senate have introduced bills (H 77 and S35) that would make it illegal for employers to use electronic monitoring tools to collect employee information without notice and consent. Employees would be protected from “termination, disciplinary action, retaliation, or other adverse employment action for refusing to follow the output of an artificial intelligence system, automated decision system, algorithm, or other similar technology” under certain conditions.
And in Connecticut, the General Assembly’s Labor and Public Employees Committee has introduced SB 1484, which would limit the use of AI to monitor employees in the workplace, require employers to inform employees of how they’re using AI, and require employers to conduct third-party impact assessments of their AI systems used in employment.
The proposal has the support of labor unions including the AFL-CIO, but in a sign of the challenges ahead, it’s already getting pushback from local business groups, including the Connecticut Business and Industry Association.
“Complying with the requirements of this bill would be extremely difficult at best, while adding significant costs on all employers to remain in compliance,” CBIA Public Policy Associate Paul Amarone wrote in opposition.
Please e-mail editors@mlex.com to contact the editorial staff regarding this story, or to submit the names of lawyers and advisers.