Dive Brief:
- California State Sen. Jerry McNerney, D-Pleasanton, announced his introduction of SB7, otherwise known as the “No Robo Bosses Act,” on March 6 to require human oversight of artificial intelligence in the workplace.
- Under SB 7, companies could not rely on automated decision-making systems, or ADS, to make employment decisions, including hiring, promotion, discipline or termination, without a human overseeing the process. The bill would also prohibit employers from allowing ADS to use personal information to predict workers’ future behaviors.
- If signed into law, the legislation would be the first of its kind in the U.S., McNerney’s office said in a press release.
Dive Insight:
McNerney, who helped set federal AI policy while previously serving as a U.S. representative, said the new legislation is not meant to prevent ADS use in the workplace but instead to establish guardrails on the technology.
“Businesses are increasingly using AI to boost efficiency and productivity in the workplace. But there are currently no safeguards to prevent machines from unjustly or illegally impacting workers’ livelihoods and working conditions,” McNerney said in a statement. “AI must remain a tool controlled by humans, not the other way around.”
The bill is co-authored by Assemblymembers Sade Elhawary, D-South Los Angeles, and Isaac Bryan, D-Los Angeles, and is sponsored by the California Federation of Labor Unions, AFL-CIO, which represents more than 1,300 unions with 2.3 million union members.
“No worker should have to answer to a robot boss when they are fearful of getting injured on the job, or when they have to go to the bathroom or leave work for an emergency,” Lorena Gonzalez, president of the California Federation of Labor Unions, said in a statement. “When it comes to decisions that most impact our jobs, our safety and our families, we need human oversight.”
Law firm Fisher Phillips said in a blog post Monday that banning predictive AI “is likely to be highly controversial” and that the bill generally “is expected to generate significant debate between labor unions and business groups.”
“Such a ban could significantly impede progress and efficiency as it ends up broadly sweeping up common uses of AI that are actually intended to help employees, including employee retention, employee satisfaction, and other similar goals,” the firm said. “Opponents will likely contend that such uses aren’t fueled by generative AI but instead simple mathematical formulas and thus not susceptible to the same concerns (hallucinations, bias, etc.) that crop up with other GenAI uses.”
The firm said states will likely take the lead on AI regulation “for the foreseeable future,” after a similarly titled bill proposed at the federal level last year, the “No Robot Bosses” Act — never made it out of committee.
“Finding no traction at the federal level, these types of bills will no doubt proliferate in the states,” the firm said.
