IA en un minutoNewsBusiness and products
Anthropic commits $150 million over three years to the US Genesis Mission
IA en un minuto newsroom · Editor: Jon Elgezabal

In 30 seconds
Anthropic will put $150 million into the Genesis Mission, a US government push to speed up science with artificial intelligence, over a three-year period. That money will bring Claude to more than 15 agencies, such as NASA, and to several hundred projects, along with training for their scientists. It extends the partnership it struck with the Department of Energy in December 2025.
Anthropic has pledged $150 million, spread across three years, to the Genesis Mission, the US federal initiative to speed up scientific and technological discovery with artificial intelligence. It announced the pledge on October 8 in Washington, DC, at "Science: A New Golden Age", a summit organized by the White House's Office of Science and Technology Policy.
The money will pay for Claude to be offered to more than 15 of the agencies taking part in the mission, among them the National Science Foundation (NSF), the National Institutes of Health (NIH) and NASA.
During those three years, several hundred of the mission's research projects will receive Claude, Claude Code (Anthropic's coding tool) and credits for its API. The company will also work alongside the agencies and the national laboratories on the government's main scientific priorities, fusion energy and quantum computing among them; it will train the scientists on those projects and give them technical support; and it will help agencies that are joining the mission to get their first projects under way. The announcement does not spell out how the money will be split or which projects will get access.
The pledge is a further step in the partnership Anthropic announced in December 2025 with the US Department of Energy (DOE) for the Genesis Mission, which at that time was the department's own initiative to apply AI and shore up the country's scientific leadership. That multi-year deal concentrated on three fields (scientific productivity, American energy dominance, and the biological and life sciences) and could reach the work of all 17 national laboratories. Anthropic says that, together with the DOE, it has since put Claude in the hands of scientists at those laboratories. Before that, working with the National Nuclear Security Administration, it had built a nuclear risk classifier, and it had deployed Claude at Lawrence Livermore National Laboratory.
So far this year, Anthropic has released Claude Science, an AI workspace that gathers the tools and packages researchers use most, and has made 10,000 Claude seats available to academic scientists either free or at a discount. It has also widened AI for Science, its program of credits for researchers on high-impact projects, and has put the Model Hardware Standard into research preview: a common specification so that AI agents can handle laboratory instruments safely.
Why it matters · analysis and opinion
Publicly funded science has become another arena of competition among the big artificial intelligence companies. Putting models in the hands of researchers at federal agencies, with credits, training and support included, builds habits, and teams that learn to work with Claude will tend to stay with it when paid contracts come along. For the government, the appeal is speeding up costly, slow fields such as fusion or quantum computing without paying up front. The practical side remains to be seen: Anthropic does not explain how the money is split or how projects are chosen, and the real value will depend on whether those scientists achieve results they would not have reached without the model.
Official source: Anthropic · Written with the help of AI: how we make the news


