States Take Legislative Action to Avoid Artificial Intelligence Discrimination in the Workplace
As of June 11, 2024, there are six states (California, Georgia, Hawaii, Illinois, and Washington) in which state legislators have proposed bills seeking greater regulation of the use of AI in making employment decisions. Employers and employees share privacy concerns and expect that any law with respect to AI would point in that direction. But while employers are still obligated by law to protect employees’ personal identification information (PII), AI or no AI, these proposed bills address AI’s unspoken fatal flaw in employment practices: discrimination and bias.
When AI first became part of the global consciousness, employers saw solutions to hiring and performance management overload and welcomed the new tool as a long-awaited answer to a question no one had known to ask. However, New York City quickly became the first to notice AI’s potential for discrimination and bias, the negative side to what appeared to be a seamless boon to employers.
New York City’s law, passed in 2021 and effective as of January 1, 2023, prohibits employers and employment agencies from using an Automated Employment Decision Tool (AEDT), a form of AI, without conducting a bias audit (impartial evaluation by an independent auditor) and notifying employees and candidates that an AEDT is being used. By law, a bias auditor does not have to be certified and can be anyone who is not associated with the employer in any way and can make reasonable and objective assessments.
Do any States Have Existing Laws?
As of now, Colorado currently has the only statewide law, passing the Colorado Artificial Intelligence Act (CAIA) on May 17, 2024. CAIA becomes effective on February 1, 2026. The law applies to “high risk artificial intelligence systems (HRAIS),” which is defined as “any systems that when deployed makes or is a substantial factor in making a consequential decision.” Employment decisions are considered consequential decisions under the law.
Under the law, Coloaction anders with 50 or more using HRAIS must develop a risk management policy and governance program, complete an annual impact assessment of the HRAIS, provide candidates and employees with a notice of the use of the HRAIS, provide candidates and employees who are subject to adverse action with a notice stating the reason for the adverse action, the opportunity to correct any personal information used, and the right to appeal the adverse action, and disclose on their website the type of HRAIS used and how they manage known or reasonably foreseeable risks of determination.
All states currently proposing bills to regulate AI, with the exception of Georgia, have included in their bills a requirement to make annual impact statements with respect to their automated decision tools (ADT). California, Hawaii, and Washington have included a required notice to employees and candidates when using ADTs to make employment decisions. California, Georgia, Hawaii, and Washington have explicitly stated that ADTs cannot be used in a discriminatory manner.
Although there is currently no federal law regarding the use of AI in employment decisions, the Biden administration has also released an executive order calling for a level playing field for the American worker in the ever-shifting world of AI.
AI Compliance Support is Available
New technology develops quickly, and lawmakers are right behind, trying to keep up. To keep up with up-to-the-minute federal, state, and local laws affecting you and your employees, from AI to the Pregnant Workers Fairness Act and beyond, myHRcounsel is here to help. With an affordable subscription for businesses of all types and sizes, myHRcounsel can help draft new workplace policies and guidance to be compliant with these laws.