Anthropic puts $200M behind real-world tests for AI job disruption

$200 million is now attached to a question most AI policy debates still treat too abstractly: what actually helps workers and institutions adapt when AI changes the shape of work? Anthropic’s Economic Futures Research Fund, dated July 22, is aimed at external research on interventions for AI-driven economic disruption.

The size and structure matter. Anthropic says it plans to prioritize projects in the $5-30 million range, with flexibility for larger well-scoped work, and will not directly fund anything below $1 million from this fund. That shifts the center of gravity away from small reports and toward field experiments, pilots, and program evaluations that can produce evidence early enough for governments and employers to use.

The fund’s five priorities show where Anthropic thinks the evidence gap is largest: how AI changes work inside firms, how people navigate transitions, how income support should adjust to displacement, how workers might gain stakes in AI-driven growth before disruption arrives, and whether public investments produce measurable results. Those are not model-card questions. They are labor-market, tax, education, and organizational-design questions.

Eligibility also keeps the program institution-heavy. Anthropic lists accredited universities and degree-granting institutions, independent research institutes, policy research organizations, and nonprofits with a track record in field experiments. Individual researchers can lead proposals through institutions, but cannot apply on a personal basis.

The broader signal is that frontier labs are beginning to fund the social instrumentation around AI, not only the technical systems. That does not settle conflicts of interest, but it does create a stream of public evidence that may become more useful than anecdotes about productivity gains or job losses. The first grant cohort will show whether the fund favors randomized trials, workplace redesign, income-support pilots, or more ambitious attempts to share AI growth with workers directly.