The Center for AI Safety has updated its Wage Loss Dashboard to include September data, and the tracker now attributes just over 117,000 job cuts nationally to AI this year, translating to an estimated $8.5 billion in annual wage losses. The dashboard is a project from CAIS that combines monthly layoff announcements with wage and population data to estimate what AI-related cuts mean for households and local economies. It presents a map of layoffs since January 2026 and a drop-down for viewing national or state-level figures.
To generate its estimates, the dashboard multiplies the number of AI-attributed job cuts by average annual salary, then divides by state population to arrive at an average loss per person. Nationally, that figure is $25 per person per year. California, with just over 27,000 layoffs attributed to AI, shows $56 per person, while Alaska, with 138 job losses, shows $14 per person. CAIS notes that state estimates are not observed AI layoffs; the project applies each month's national share of cuts citing AI to a state's announced job cuts.
The dashboard draws on monthly reporting from Challenger, Gray, and Christmas, annual mean wage data from the Bureau of Labor Statistics, and population estimates from the US Census Bureau. In 2025, Challenger, Gray, and Christmas identified over 54,000 job cuts that companies publicly linked to AI, equal to one in 20 layoffs that year. CAIS said that while job loss from AI has drawn significant coverage, quantifying the effect on the average person's finances is difficult, and the dashboard is meant to make the monetary costs visible to workers, communities, and policymakers.
CAIS describes its methodology as a moderate standard: it does not require an executive to say a team was replaced by AI, a choice the researchers said captures the broad pattern of AI-justified workforce reductions while excluding unrelated cuts. The project points to ripple effects through grocery stores, small businesses, and rent, noting that one person's layoff reduces income for the businesses they support. States with large tech and media sectors are seeing greater impact, according to CAIS, which also said it does not intend to shame any industry or company for its AI adoption choices.
The dashboard also covers how well AI completes remote freelance tasks, using the Remote Labor Index benchmark that CAIS co-developed to test agents on real-world work such as web development and video production. When RLI launched, the best AI agent completed 2.5% of projects at professional quality, a figure that has since risen to 15.8%, with that percentage doubling every 3.7 months, according to the researchers. Claude Fable 5 ranks first on the benchmark, followed by OpenAI's GPT-5.5. RLI measures what models can do, not what they are actively taking on.
CAIS lists several caveats: the dashboard excludes jobs created by AI, productivity gains, re-employment, and secondary economic effects, partly because experts lack consensus on whether AI will stimulate the economy or outpace displacement. It also notes that AI-attributed layoffs are murky because there is no single definition of what that means and companies are not fully transparent, making it difficult to prove a layoff resulted directly from AI. Separately, Jobloss.ai, run by the Alliance for Secure AI, logs layoffs of at least 100 jobs at once that come with some mention of AI or terms such as automation and productivity. Recent laws in a few states, including California, aim to ensure firing and other disciplinary decisions are not fully automated.
More AI news from TechManNews.
