Anthropic's $200M Research Fund: Five Questions It's Paying to Answer About AI and Jobs
Anthropic has published a five-priority research agenda for its $200M Economic Futures Research Fund, targeting AI-driven worker displacement.
Anthropic has published the research agenda for its Economic Futures Research Fund, committing $200 million to external research on how society can prepare for AI-driven economic disruption. This is not a vague corporate responsibility pledge. The company has named five specific research priorities, set grant sizes between $1 million and $30 million, and said it is looking for large-scale randomised controlled trials and ambitious pilots, not incremental studies.
The announcement sits alongside a broader $350 million commitment that includes a $150 million fellowship placing early-career workers inside nonprofits for 12 months. But this post is about the research fund specifically, because the five-priority agenda is worth understanding on its own terms.
What the Five Priorities Actually Are
Anthropic has been explicit about what it wants to fund. Here is each priority and what it is trying to find out.
1. How AI affects workers at the firm and workplace level
Most existing evidence on AI in the workplace is observational and short-term. Anthropic wants field experiments that can answer harder questions: which patterns of human-AI collaboration actually develop worker expertise over time, how organisational design choices affect who captures the productivity gains, and what difference it makes when workers have a voice in those design choices. This is research that companies generally do not fund themselves, because the findings might be uncomfortable.
2. Helping people navigate AI-driven transitions
The fund will support tests of what Anthropic calls “ambitious mobility instruments,” including paid leave tied to retraining programmes and portable benefits that move with workers across employers rather than staying with a single job. There is already evidence that some sectoral training programmes work well. The open question is whether they can scale quickly to a much broader population. Anthropic specifically mentions the idea of a large-scale “fire drill,” scaling a promising programme rapidly across job seekers in a given state to see what breaks and what holds.
3. Modernising income support
Most unemployment insurance systems, including the US system, were designed around the assumption that job loss is temporary. Anthropic’s Economic Policy Framework acknowledges that AI-driven displacement may be neither temporary nor narrowly targeted. The fund will research what updated income support looks like for a different kind of disruption. CEO Dario Amodei’s accompanying essay outlines a graduated policy menu covering scenarios where national unemployment reaches 5%, 10%, and beyond, including long-term income support and, in the most severe scenario, universal basic income financed by taxes on AI companies or capital gains.
4. Building worker stakes in AI-driven growth
This is perhaps the most forward-looking priority. Anthropic wants to explore mechanisms that give workers a share of the economic value AI generates, before widespread disruption arrives rather than after. Capital accounts and equity-sharing models are mentioned. The logic is that it is easier to build these structures during growth than to retrofit them during a crisis.
5. Generating evidence on public investment
The fifth priority is about community-facing public services: what kinds of public investment actually help people and communities adapt, and what does the evidence say about return on those investments? This is research that is hard to do without sustained funding, because the relevant outcomes take years to materialise.
Who Can Apply, and What Anthropic Is Looking For
Eligible organisations include accredited universities, independent research institutes, policy research organisations, and nonprofits with a track record of running field experiments at scale. Grants range from $1 million to $30 million, with most projects expected to fall in the $5 million to $30 million range.
This is a deliberate shift from Anthropic’s earlier Economic Futures Programme, which offered grants between $10,000 and $50,000. The company says it learned that managing many small grants simultaneously is hard to scale, and that its comparative advantage is funding large bets that other funders would not take on.
Anthropic is also asking for something unusual: research partners willing to share findings publicly at key milestones, so that policymakers can act on results while they are still timely. That condition matters. It means Anthropic is at least structurally committed to findings reaching the public domain, even if those findings are inconvenient.
What This Signals About Anthropic’s Positioning
It would be easy to read a $200 million research fund as reputation management from a company preparing for a public listing. That reading is not entirely wrong. Anthropic is a private firm developing some of the most capable AI systems in the world, and publicly funding research into the harms those systems might cause is good optics.
But the specificity of the agenda makes it harder to dismiss. Firm-level field experiments on who captures productivity gains, portable benefits that challenge the employer-tied benefits model, equity-sharing mechanisms built before disruption hits: these are not safe, comfortable research topics. They imply findings that could point toward structural changes that the AI industry would not necessarily welcome.
The broader context sharpens the picture. Entry-level tech hiring dropped an estimated 30 to 50 percent in 2025 compared to 2023 levels, and nearly 78,600 tech workers lost jobs in the first quarter of 2026 alone, with close to half of those cuts attributed to AI automation. Anthropic is not the only company to have published a jobs framework in this period. By its own count, virtually every major AI company put forward some version of a plan between April and mid-June 2026.
What distinguishes Anthropic’s approach is the emphasis on generating evidence rather than advocating for a specific policy outcome. The Economic Policy Framework acknowledges that the most promising solutions for an unprecedented level of disruption may be ones nobody has tried yet. Funding large-scale pilots and randomised trials is a reasonable response to that uncertainty, provided the results are actually used.
What This Means for You
The honest answer depends on who you are.
If you are a researcher or at a policy institution, this fund is worth paying attention to. The grant sizes are substantial, the research questions are genuinely open, and Anthropic says it will consider ambitious proposals outside the five stated priorities if they are calibrated to the scale of the problem. Applications can be submitted through a request for proposals.
If you are an early-career professional, the accompanying $150 million Claude Corps fellowship is more immediately relevant. The first cohort of 100 fellows will be placed at nonprofits for 12 months at $85,000 per year plus benefits, with applications closing 17 July 2026.
If you are a worker concerned about AI and job security, the research priorities speak directly to your situation, even if the findings are years away. The questions being funded, particularly around who captures AI productivity gains and whether income support can be modernised for persistent rather than temporary displacement, are the right questions to be asking. Whether the answers translate into policy is a separate problem, and one that will depend on political will as much as evidence.
The real test for this fund is whether it produces findings that occasionally make Anthropic uncomfortable, and whether the company publishes them anyway. That is worth watching.