The issue of regulating AI has been circulating amongst federal, state, and local legislation. With this, the California Civil Rights Department has proposed a few regulations to protect against employment discrimination in automated decision-making systems.
Proposed AI Regulations
To protect against discriminatory acts during the employment process, the CRD has proposed clarifying that if an automated decision-making system harms an applicant or employee, it’s a violation. Employers must also maintain all employment records, including data from AI decision-making systems, for at least four years.
Additionally, CRD proposes that AI decision-making does not replace individualized assessment when considering an applicant’s criminal history. Third parties are not allowed to aid in employment discrimination using automated decision-making systems. The CRD also proposes to add definitions for key terms used, such as “automated decision system,” “proxy,” and “adverse impact.”
Although these regulations are simply proposed, it’s still difficult to see how they could affect other pending legislation. An example of this is Assembly Bill 2930, which requires notification and bias audits when using AI in employment.