We recently wrote about Colorado’s historic law aimed at protecting, among others, employees and employment applicants from harm arising out of the use of artificial intelligence (AI) systems. Although Colorado is the first state to pass legislation addressing AI-based discrimination, similar bills have been proposed in at least six other states as well as at the federal level, with a recent Executive Order discussing a wide range of issues arising from the private-sector use of AI systems, including discrimination in the employment context. All US employers should take note of this trend and prepare for the possibility of new compliance obligations resulting from employer use of AI systems.

States with proposed legislation addressing algorithmic discrimination

In 2024, legislators in California, Georgia, Hawaii, Illinois, and Washington proposed bills aimed at regulating the use of AI systems to make, or to assist an employer in making, employment decisions. More specifically, these bills seek to mitigate the risk of algorithmic discrimination arising from an employer’s use of an AI system.

 “Algorithmic discrimination” means any condition in which the use of an AI system results in differential treatment or impact that disfavors an individual on the basis of protected characteristics (e.g., age, color, ethnicity, disability, national origin, race, religion, veteran status, and sex).

Each of the proposed bills would impose similar obligations onto employers that use an AI system or an automated decision tool (ADT) in making employment-related decisions. ADT means a system that uses AI and has been specifically developed to make, or contribute to making, consequential decisions, including employment decisions.

In general, these proposed bills impose a duty of reasonable care on employers to mitigate and assess the risk of algorithmic discrimination arising from their use of AI systems. There are significant affirmative reporting requirements, including direct notifications to individuals who are the subject of a decision made by an AI system. In some cases, the bills also provide individuals with the chance to correct data input into the AI system and appeal adverse consequential decisions, which may require human review. The specifics of each state’s proposed bill are discussed further below.

California: AB 2930

• The bill would (i) prohibit employers from using ADTs in a discriminatory manner; (ii) require employers to perform annual impact assessments of its use of an ADT; and (iii) require employers to provide notice to applicants and employees that are subject to the use of an ADT.
• The law provides a narrow exception for employers with fewer than 25 employees.
• If passed, the bill will go into effect on January 1, 2026.

Georgia: H.B. 890

• The bill would (i) expand existing anti-discrimination laws to prohibit discrimination resulting from the use of or reliance upon AI or ADTs, and (ii) preclude employers from relying upon the use of AI or ADTs as a defense to an allegation of discrimination.
• The proposed bill does not provide an effective date.

Hawaii: H.B. 1607

• The bill would (i) prohibit employers from using ADTs in a discriminatory manner; (ii) require employers to perform annual impact assessments of its use of an ADT; and (iii) require employers to provide notice to applicants and employees that are subject to the use of an ADT.
• The law provides a narrow exception for employers with fewer than 50 employees.
• The proposed bill does not provide an effective date.

Illinois: H.B. 5322

• The Illinois Commercial Algorithmic Impact Assessments Act would require annual impact assessments by employers that use AI systems.
• The law provides a narrow exception for employers with fewer than 50 employees.
• If passed, the bill will go into effect on January 1, 2026.

Washington: H.B. 1951

• The bill would (i) prohibit employers from using ADTs in a discriminatory manner; (ii) require employers to perform annual impact assessments of its use of an ADT; and (iii) require employers to provide notice to applicants and employees that are subject to the use of an ADT.
• The law provides a narrow exception for employers with fewer than 50 employees.
• If passed, the bill will go into effect on January 1, 2025.

In California and Washington, legislation aimed at addressing algorithmic discrimination previously failed in 2023. However, in both states, the currently-proposed bills now carry the support of several large, influential technology-based employers.

What federal regulations addressing algorithmic discrimination could contain

No federal legislation addressing AI-based discrimination in employment has been proposed at this time. That said, however, the President’s recent Executive Order indicates that issues arising from the use of AI systems also are being considered at the federal level. When Colorado Governor Jared Polis signed the Colorado Artificial Intelligence Act (“CAIA”) into law, he expressed his opinion that “the important work of protecting consumers from discrimination” arising from the use of AI systems “is better considered and applied by the federal government[.]”[1] Governor Polis also suggested that there is a chance that the federal government preempts CAIA “with a needed cohesive federal approach.”[2]

In his Executive Order, President Biden directs the federal government to “enforce existing consumer protection laws and principles and enact appropriate safeguards against…unintended bias, discrimination…and other harms from AI.” In addition, the Assistant Attorney General in charge of the Department of Justice’s Civil Rights Division was required to meet with the head of the federal government’s civil rights office to discuss comprehensive use of their respective authorities to prevent and address algorithmic discrimination in the use of AI systems.

Should the federal government pass regulations on employers’ use of AI systems, the requirements will likely mirror legislation addressing the same issue at a state level. Thus, federal regulation could impose on employers affirmative reporting requirements related to the use of an AI system and a duty of reasonable care to mitigate the risk of algorithmic discrimination when using AI systems to make employment decisions.

What all US employers need to know

Although state legislatures and the federal government only recently began to tackle issues arising from the use of AI, including by employers to make employment related decisions, the sheer number of proposals addressing the issue suggests that the trend towards regulation of AI systems will continue moving forward and moving fast. Beyond the scope of this blog, legislators across the nation have also proposed various bills addressing other issues arising from the use of AI, such as pornographic deepfakes and the use of AI chat bots.

Colorado still is the only US jurisdiction that currently regulates an employer’s use of AI systems. However, employers outside of Colorado should take note of how regulation at the federal level and at the state level could impact their use of an AI system or ADTs.

Our team will continue to monitor these and other state employment law developments. 

[1] SB24-205 Signing Statement.pdf – Google Drive

[2] Id.