Ongoing Developments As to the Use of AI in Hiring

Authors: Summer Associate Maggie Newhart & Associate Ula Chun

Introduction

As more employers rely on artificial intelligence (AI) tools to streamline hiring, the legal risks increase. While it appears AI can offer efficiency and consistency in recruitment, issues as to bias and fairness are presented.

Federal Workplace Laws and AI

The federal approach to AI in the workplace is in flux. Earlier this year, President Trump rescinded President Biden’s Executive Order 14110, which was designed to address the risks of AI use, including discrimination, bias, and compromised safety. Also withdrawn was the Equal Employment Opportunity Commission’s guidance touching on potential AI-related violations of Title VII of the Civil Rights Act of 1964 and the Americans with Disabilities Act (ADA).

Nevertheless, companies remain vulnerable to discrimination suits based on the use of AI in workplace decisions. For instance, in July 2024, the U.S. District Court for the Northern District of California denied a motion to dismiss the claim that Workday, an AI-powered hiring and HR software, repeatedly discriminated against plaintiff Derek Mobley by screening him out of jobs due to race, age, and disability. The Court held that Workday could be directly liable for employment discrimination under the theory that an AI service provider can be an “agent” of the employer.  Mobley v. Workday, Inc., 3:23-cv-00770. The Court also granted preliminary conditional certification of a collective action under the Age Discrimination in Employment Act.

California Civil Rights Council Regulations

Meanwhile, in the absence of clear federal guidelines, California is one of several states that has enacted a series of laws and regulations that establish new standards for AI in the workplace. In June 2025, the California Civil Rights Council secured approval for regulations on employers’ use of AI, algorithms and other Automated Decision Systems (“ADS”). These regulations go into effect on October 1, 2025. Under the regulations, it is unlawful to use ADS if doing so has a discriminatory effect, unless doing so is justified by a permissible defense such a business necessity.

More specifically, there are four particular pitfalls employers must steer clear of to avoid liability under these regulations.

• First, relevant to any claim or defense about ADS will be the evidence – or lack of evidence – of anti-bias testing or similar proactive efforts to avoid unlawful discrimination.
• Second, parties must take care using ADS to screen out applicants based on their schedule or availability because this can unfairly impact people with disabilities or who may require time off for religious observance.
• Third, parties must take care using ADS to screen, interview, or measure applicants and should not use AI to measure personal traits. Tools that assess an applicant’s skill, reaction time, physical dexterity, facial expressions, or tone of voice may put some people at a disadvantage, e.g., due to their disabilities, race, gender, or national origin. Accommodations or other proactive steps may be necessary.
• Fourth, the regulations make clear that employers are legally accountable for the third-party AI vendors they use to exercise employment functions (such as recruitment, screening, or hiring) because such vendors are treated as agents of the employer. Further, a vendor could also be deemed to be an employer under the rules.

In addition to these new rules, the California Civil Rights Council has outlined new recordkeeping requirements. Employers must retain ADS-related records for at least four years.

California State Legislative Efforts

There are two bills related to AI and employment decision-making pending approval in California.

AB 1018, awaiting Senate approval, would require that, before any consequential decision is made or facilitated by ADS, the subject of the decision is provided detailed notice, mandatory opt-outs, and the right to appeal of the decision. The bill requires long-term recordkeeping and ADS performance evaluations including disparate treatment or impact analysis and impact assessments by auditors. Penalties of up to $25,000 per violation would be available.

SB 7, or the “No Robo Bosses” Act, would require written notice that ADS is used in employment-related decisions, both before and after use thereof. It also would preclude AI tools from collecting certain information or limiting its use in certain decisions. Workers affected by certain decisions made by ADS would have the right to appeal the decision.  The bill also mandates human oversight and 30 days’ notice before using ADS in employment-related decisions. It awaits Assembly approval.

Mitigating Legal Risks – Employer Checklist

Given the evolving legal standards around AI in employment decisions, employers should consider taking the following steps to minimize legal risk and ensure compliance with new regulations.

1. Develop an AI policy to include in updated employee handbooks
   1. Define allowed uses of AI in hiring practices.
   2. Outline transparency, nondiscrimination, and accountability goals.
2. Screen AI Vendors
   1. Find out how the system works, where it gets its data, and what it considers in making decisions.
   2. Avoid systems that lack transparency (“black box AI”).
   3. Understand what anti-bias testing or other proactive efforts have occurred.
3. Provide Staff Trainings
   1. Educate HR and hiring staff on how to use AI tools appropriately and the risks that may arise as a result of irresponsible AI usage.
   2. Inform staff that employers remain legally responsible for third-party AI tools that are used in the hiring process.
4. Keep Humans as the final decisionmakers
   1. Emphasize that AI is only meant to assist in the hiring process.
   2. Require that final decisions be reviewed and approved by a human.
   3. Keep precise records of decision-making processes in case of future litigation.
5. Periodically Audit AI Systems
   1. Keep records and review outputs for disparate impact on protected classes.
6. Give Notice and Get Consent
   1. Disclose use of AI when legally required, and consider voluntary disclosures even when not required as a way to build trust, show good faith and potentially avoid claims.

By taking these steps, employers can protect begin to themselves from expensive litigation and penalties, while supporting the inevitable integration of AI into their hiring practices.