California introduces its inaugural AI job-loss monitoring system.

      California has created the first state tool to monitor the impact of AI on job loss. The initial findings indicate no widespread layoffs yet, but there are warning signals emerging in the Bay Area and among workers with college degrees.

      The debate continues over whether AI is responsible for job losses, but concrete data has been scarce. California has now developed a tool to gather some insights. The state has introduced what it describes as a pioneering system for tracking AI-related job losses in real-time.

      This dashboard is named the California AI-Unemployment Tracker, or CAIT. Governor Gavin Newsom announced it on Thursday, referring to it as an early warning system. His office collaborated with the California Policy Lab, a nonpartisan research institution at the University of California, and the state Employment Development Department to create it.

      The innovative aspect lies in the methodology. It analyzes California’s monthly unemployment insurance claims and categorizes each claim based on the worker’s job exposure to AI. By monitoring that proportion over time, trends may emerge before they become mainstream news. The data is updated monthly and is accessible for public download.

      Current data findings show a sense of calm, albeit with a caveat. Researchers found no evidence of a spike in AI-related layoffs statewide. Since the introduction of ChatGPT in late 2022, the proportion of claims from AI-exposed workers has not significantly increased.

      However, a deeper analysis reveals shifts within specific groups. After the release of ChatGPT-3.5, claims from college-educated workers in high-exposure jobs surged and have remained elevated through May 2026, while those in low-exposure positions did not experience similar changes.

      Geographically, a similar pattern is observed. Claims increased significantly among AI-exposed workers in the San Francisco Bay Area and were notably high in tech sectors such as information and professional services. The data didn’t show significant variations based on race, gender, or age.

      “Currently, there is no evidence of widespread AI-induced layoffs in California’s labor market,” stated Ben Hyman, a senior researcher at the lab and co-author. “However, there are notable trends in specific regions like the Bay Area, within tech-heavy sectors, and among highly AI-exposed workers with college degrees.”

      The caveats are important to note. The lab is clear that the tool cannot establish that AI led to any particular layoff; no dataset can accomplish that. The tracker serves as an indicator, not definitive proof, as researchers emphasize.

      There are limitations as well; unemployment claims do not account for gig workers, self-employed individuals, or those who don't file claims. Job titles are self-reported without verification, and the researchers omit pandemic data to prevent it from overshadowing current trends. The honest assessment is that this is a preliminary tool, not a final judgment.

      The exposure score is based on two measurements. One, developed by OpenAI and academic researchers, assesses whether an AI model could perform at least half of a job's tasks. The other, derived from Anthropic’s economic index, tracks how frequently workers utilize AI in their roles. Together, these metrics provide a more comprehensive perspective than either could alone.

      The differentiation is intuitive upon examining examples. The tracker rates customer service representatives and software developers as highly exposed, while it places heavy-goods drivers and nursing assistants towards the bottom. “This new tracker transitions speculation into evidence,” said Till von Wachter, the UCLA economist who co-led the initiative, “and gives us greater clarity on the changes taking place.”

      The timing of the release is significant. Newsom is widely anticipated to run for president in 2028. Politicians across both parties are eager to appear proactive regarding AI and employment as voters express concerns over costs and observe data centers increasing local utility bills.

      California has a particular incentive to act, as it has the highest unemployment rate in the nation and is home to many companies developing advanced AI technologies. This positions it as both a hub for innovation and a crucial testing ground for its consequences.

      Public sentiment has shifted as well. American attitudes towards AI have become more negative, coinciding with increased adoption. Approximately one in five U.S. workers now uses AI in their jobs, showing a significant increase from the previous year. The same tools that people are beginning to depend on are also those they fear.

      California is not the first state to explore this issue but is the most ambitious. New York amended its layoff-notice regulations in 2025 to highlight cuts associated with AI, and Connecticut enacted a similar measure last month. The flaw is evident, as it relies on companies to self-report. Out of more than 160 New York companies that announced mass layoffs, none attributed those layoffs to AI.

      The underlying anxiety driving this action is not unfounded. American automakers have already cited AI as a reason for job cuts within their white-collar workforce. California is also engaged in legal battles over alleged bias in AI hiring practices. Meanwhile, public trust continues to decline.

      The more challenging question is how the