FAR pipeline paper published: in a combinatorics pilot, human mathematicians verified all 15 selected AI-produced results as correct
Anthropic Opens 10,000 Free and Discounted Claude Seats for Scientists
Subsidized access to frontier AI tools lowers a cost barrier that had previously limited academic and nonprofit adoption. The biosafety carve-outs show Anthropic maintaining restrictions on dual-use biological queries even as it broadens science access.
The full picture
Anthropic is offering 10,000 seats on a Claude Team plan for scientists, with free standard seats and premium seats at $15 per month (an 80% discount) for one year. Principal investigators at academic or nonprofit research institutions must verify to qualify and can then add lab members. Premium seats carry 5x usage limits. The AI for Science program is expanding beyond biological sciences, with up to $50,000 in credits per project for compute-heavy work such as research on the Riemann zeta function and protein design. Biology and chemistry researchers are currently limited to Opus-class models; Anthropic is working with the US government on a life sciences access program for Mythos-class models. Claude Fable models will continue to block professional biology and drug development queries due to dual-use risks. Anthropic plans to expand beyond the initial 10,000 seats in coming months.
How it developed
Anthropic announces 10,000 free and discounted Claude Team seats for scientists, expanding AI for Science credits to fields beyond biology.
Cambridge and Hong Kong University paper argues autonomous AI agents discover mathematical results fixed scaffolding would not
Anthropic mathematician claims to have solved the S^6 complex structure problem (open since 1948) using Claude
Sources
Want this in your inbox?
I send one email each morning with the stories that moved. If you would rather just read here, that works too.
Subscribe free