On June 27, 2025, Anthropic announced the Economic Futures Program, a research and policy initiative aimed squarely at AI’s economic disruption. The stated mission: understand how AI is reshaping the way people work, and surface proposals for preparing for the shift. It is the first formal, long-running program from a major AI lab dedicated to the technology’s impact on the labor market.
The timing is pointed. In May, Anthropic CEO Dario Amodei had publicly warned that AI could eliminate half of all entry-level white-collar jobs and push unemployment as high as 20% within one to five years. The lab that delivered the loudest warning then built its own research program — TechCrunch’s headline put it as launching “as job losses loom.”
The program’s three pillars
Per the official announcement, the program rests on three pillars. First, Research Grants: funding, API credits, and partnerships for researchers studying AI’s effects on labor, productivity, and value creation. Second, Evidence-Based Policy: convening forums where researchers, policymakers, and practitioners develop and evaluate policy proposals for AI’s economic impacts, grounded in real-world data. Third, Economic Measurement and Data: scaling the existing Economic Index into longitudinal datasets tracking AI’s economic usage and impact, with regular public data and research releases. The pillars build on each other by design — better data enables better research, which informs better policy.
The backdrop: Amodei’s warning on jobs
To read the program, start with Amodei’s May 2025 comments predicting that AI could wipe out half of entry-level white-collar positions and push unemployment to 20% in one to five years — one of the starkest public forecasts about its own technology from any lab CEO. TechCrunch framed the launch against that warning: the loudest alarm came from inside the industry, and society still has no consistent framework for responding.
Built on the Economic Index
The program does not start from zero. Its data backbone is the Anthropic Economic Index, launched in February 2025, which aggregates and anonymizes usage data from Claude to show which occupations and tasks AI is actually used for, and publishes that data openly. TechCrunch flagged the contrast: most competitors lock this kind of usage data behind privacy policies, while Anthropic releases aggregated data for researchers. The third pillar’s job is to turn that snapshot into a longitudinal dataset capable of showing trends over time.
Open questions: can it stay neutral
Sarah Heck, Anthropic’s head of policy programs and partnerships, deliberately framed the program’s posture as fact-finding first. “Everybody’s asking questions about what are the economic impacts [of AI], both positive and negative,” she told TechCrunch. “It’s really important to root these conversations in evidence and not have predetermined outcomes.” The key goal, she said, is to figure out what is actually happening — if there is job loss, convene thinkers on mitigation; if there is huge GDP expansion, convene policymakers on that instead.
The skepticism is easy to anticipate: a company whose revenue grows with AI adoption is funding research into AI’s social impact, so will conclusions skew optimistic? Heck’s answer is that independent researchers make the judgments, with Anthropic supplying data and resources rather than conclusions. The program’s credibility will rest on how much data it actually opens, and on handling of results that cut against its business interests. For researchers tracking AI’s labor-market effects, the grants and API credits are real resources; for industry watchers, the metric to watch is whether these datasets become the standard citation in policy debates.
First steps and what to watch
Per the announcement, the program’s first move is funding independent researchers and experts through Economic Futures Research Awards to study AI’s economic impacts and propose policy responses. For mid-2025 observers, three things are worth tracking. First, the release cadence and granularity of the longitudinal datasets — that determines their real value to academic research and policy work. Second, whether the forums’ policy positions ever diverge from Anthropic’s commercial interests, and how the company handles those moments. Third, whether other labs follow with comparable programs. If the Economic Index data becomes a standard source for labor economists, Anthropic gains an asymmetrical seat in defining how AI’s economic impact is measured; if the data releases stall, the program is public relations. The test will be the actual release record over the coming quarters, not the announcement copy.
Sources
- Anthropic: Introducing the Anthropic Economic Futures Program
- TechCrunch: As job losses loom, Anthropic launches program to track AI’s economic fallout
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.