Oakland Union Wins AI Job Protections as US Public Workers Bargain Over Automation

Members of the International Federation of Professional and Technical Engineers Local 21 voted on October 22 to authorise the strike, with 94.1% in favour. They were concerned that the city would bring in AI systems that increase surveillance of staff and assess their performance inaccurately.

The Guardian reported on the Oakland deal and similar disputes elsewhere. Its interviews show that US public-sector unions are now using contract negotiations to set terms for how governments adopt AI. Agencies are adopting the technology quickly. Nearly 70% of state and local government agencies said they had adopted or were implementing generative AI, according to an Ernst and Young survey conducted in July and August 2026. For vendors selling AI into government, this means unions may now negotiate over how their products are used.

"We just want the city administration to really be intentional about how it's engaging with AI products," Julian Ware, an Oakland spatial data administrator and Local 21 member, told the Guardian.

Labour leaders said they were acting because lawmakers had not. "We're using our contract negotiations to institute guardrails that policymakers won't," said Liz Shuler, president of the AFL-CIO and founder of its Technology Institute.

Pennsylvania offers another example. Last year it became the first state to run a pilot with OpenAI and encouraged employees such as unemployment claims staff and social workers to test ChatGPT Enterprise for time savings. Steve Catanese, president of Service Employees International Union Local 668, which represents more than 10,000 public-sector workers in the state, said the pilot came with no guarantees for staff. "With generative AI, workers are building their own gallows sometimes," he said. The union then pressed Governor Josh Shapiro to adopt a policy guaranteeing that the state's use of AI would include employee input and supervision and would not be used to discipline workers.

Transit workers describe a similar fight over automation. "Our members know that automation is the ultimate threat to their jobs," said John Samuelsen, president of the Transport Workers Union and a New York City subway track inspector.

Kung Feng, a researcher at the University of California, Berkeley's Labor Center, co-authored an October report on what researchers call bossware. The term covers AI tools that rank workers for hiring, firing and promotion, track their location and activity, scan biometrics for timekeeping or use analytics to try to predict emotions. Feng said unions' demands usually include advance notice before a new technology is introduced, information about how it works and the right to negotiate over its use. He noted that the Communications Workers of America recommended AI protections in 2023 and won some of them for California workers in October 2026.

Lorena Gonzalez, president of the California Federation of Labor Unions, said some automation, such as vehicle safety systems in public transit, could help workers, but only if people stay in charge. "If we're using a computer or AI, it has to be controlled by a human being, not the other way around," she said.

The Oakland agreement still needs final approval. If it is approved, it will be one of the first public-sector contracts in the US that explicitly protects jobs from AI displacement.