New California rules on artificial intelligence used to watch and manage workers are about to take effect, putting the state further than any other in regulating so-called bossware, and an employment lawyer warns that many businesses may not realize the software they already run falls under them, KQED reported Wednesday.
At the end of September, Gov. Gavin Newsom signed five bills backed by organized labor that regulate AI in the workplace. Two stand out:
- AB 1883 limits employers' use of AI-driven workplace surveillance tools to read workers' emotions or collect their neural data.
- AB 1979 bars health facilities from letting AI by itself handle work reserved for licensed professionals, and requires them to protect clinicians' independent judgment when AI informs patient care.
A nurse's sentiment score
KQED's report centers on a Kaiser Permanente telephone triage nurse in the Central Valley who said she found a "sentiment score" attached to her recorded calls after returning from maternity leave, following a 2024 pilot that nurses had been told was paused. She said she could not tell what scale the numbers used.
A Kaiser spokesperson told KQED the company does not "currently" use AI to recognize, predict or infer nurses' emotional states and is reviewing what AB 1883 means for its operations. Kaiser did not answer KQED's questions about whether its call platform produces sentiment scores.
Anthony Zaller, a lawyer who represents employers, told KQED that scoring whether a worker said the right words is acceptable but scoring what the worker was feeling is now off limits. He said companies often do not realize how deeply AI is built into the products they buy, and that they cannot avoid responsibility under existing law because a decision ran through software.
