The AI Safety Net Is Being Built After the Layoffs
AP reported this week that RAISE US, a new bipartisan nonprofit led by former Commerce Secretary Gina Raimondo and former Indiana Gov. Eric Holcomb, launched with $500 million in backing from OpenAI, Anthropic, Amazon, Microsoft, IBM, and Bank of America — the same ecosystem accelerating AI adoption in U.S. workplaces. Pilot programs will begin in Arkansas, Maryland, Utah, and Connecticut, focused on preparing workers for AI-driven labor disruption that forecasts from Boston Consulting Group and Goldman Sachs project could automate up to 25 million U.S. jobs within five years.
The initiative acknowledges a basic structural problem: the current workforce retraining system was not built for a technology that can affect white-collar, administrative, creative, and technical work simultaneously. Traditional programs move slowly, operate unevenly by state, and tend to respond to past labor shocks rather than anticipate the next one. RAISE US is being positioned as a faster, more coordinated infrastructure for that transition.
The companies funding the safety net are the same ones building the tools producing the disruption. That does not automatically make the effort illegitimate, but it does create a material tension between worker protection and employer interest in managing the pace of AI adoption. A retraining program funded by the companies displacing workers is not the same thing as a retraining program designed from the worker’s position.
This is not the beginning of the AI labor story. It is the third act. First came the tools. Then came the layoffs. Now comes the infrastructure that frames the disruption as a transition workers can absorb — rather than a shift that workers should have had a role in shaping from the start. Whether RAISE US produces genuine workforce security or a more managed version of the same economic pressure depends on implementation details that its funders will have significant influence over.
