ZDNET’s key takeaways
- AI is playing a bigger role in hiring and firing decisions.
- New California legislation protects workers from fully automated personnel decisions.
- It could set the tone for other state or national legislation.
AI use at work is on the rise, touching everything from everyday tasks to hiring and performance reviews. An April survey by MyPerfectResume found that of 1,000 HR professionals, 73% said they use AI to make decisions and 52% said they use AI to inform “restructuring and role planning.”
Hiring managers are also turning to AI for deeper behavioral insights when screening candidates: More than half of respondents (51%) use it to flag “risky” candidates. Only 26% of respondents said they do not use AI to make decisions about layoffs.
Also: How to beat the AI algorithm and get the job of your dreams
California Governor Gavin Newsom recently signed new legislation protecting workers from certain AI-powered decision makers in the workplace. The bill covers these issues from several angles — here’s what the regulation tries to address, and how it could set the tone for future state or federal legislation.
AI-powered burning
SB 947, somewhat cheekily called the “No Robo Bosses Act,” prohibits employers from outsourcing disciplinary and termination decisions entirely to AI, referred to in the bill as automated decision-making systems (ADS). The bill requires employers to review all AI decisions regarding employee discipline or dismissal, as well as provide the employee with a description of the reasons for the decision. That includes all data used to make the decision, including personnel files, evaluations, work product, peer reviews, and “witness interviews.”
“If an employer cannot confirm the ADS output or the human reviewer has concluded that the ADS output is inaccurate, incomplete or misleading, the employer shall not use the ADS output to make a disciplinary or termination decision,” the bill states.
Also: AI just broke your career ladder – here are 6 new ways to the top
The law, which takes effect on July 1, 2027, allows affected employees to file complaints with the California Labor Commissioner if they believe that this process was not properly followed in a firing decision. It does not fully prohibit the use of automated systems for these determinations. It also requires employers to clarify whether “a mass layoff, relocation, or termination is caused by an AI system,” according to a brief posted on the governor’s website.
While AI regulation at the federal level remains less specific, sprinkled in voluntary agreements with labs, and arguably focused on distractions, states have risen to the challenge of creating initial protections for workers. However, when it comes to AI in hiring and other workplace processes, some states like Colorado and Connecticut have chosen to mandate disclosure of AI use rather than actual intervention.
California and New York have tended to lead the charge on stricter laws, though Illinois passed a landmark bill this summer that will become the first state to require third-party audits of border labs.
Also: 73% of tech job listings now require AI skills: How to show yours
Lily Li, a lawyer specializing in AI and founder of Metaverse Law, told ZDNET that she sees California’s latest law as a precedent for other states. But she also expects companies to find loopholes in the law-required human review process.
“We could see employers engaging in check-the-box assessments rather than meaningful human review to avoid the operational costs of an additional layer of review,” she said. “The ability to ask an employee for an explanation is a helpful check against this, though.”
With this concern, it is possible that the onus to address unaddressed AI-driven fires will fall on complainants who contest their own fires in court. Still, protections like these will remain patchwork across the country until a version of them is adopted at the federal level.