Anthropic and Citrini Use Claude to Design Scorpion Neurotoxin Binders
Anthropic and Citrini Research used an internal Claude model to design protein minibinders targeting scorpion neurotoxins, yielding 17 lab-confirmed binders from 211 tested designs.
Proof1 independent outlet1 community thread
Key points
- Working within Claude Science across 64 hours, the internal model operated mainly autonomously with four prompt follow-ups, analyzing research data, generating figures, and drafting a report.
- Claude generated 318,484 protein designs, selecting 180 for testing alongside 31 additional designs added by the team.
- Contract research organizations Adaptyv Bio and Twist Bioscience experimentally tested all 211 designs, finding 17 that met criteria for binding to one of four human-dangerous scorpion neurotoxins.
- Anthropic emphasized that it has not confirmed cellular venom neutralization or produced an antivenom, while external reports noted the model did not create an actual drug.
- All 211 protein designs and their experimental data are being released as a public resource.
Why it matters
- The project demonstrates that an AI model can design credible binders for hard-to-target toxins within days, providing publicly accessible design data that could accelerate discovery for neglected diseases.
Sources (3)
- [1]How Claude designed protein binders against scorpion neurotoxinsAnthropic Research · Oct 20, 12:00 AM
Summary: As part of our work experimenting with how Claude can help accelerate drug discovery, we asked an internal research model to design minibinders (small proteins that attach tightly to a target) against scorpion neurotoxins. Claude worked from our prompt and four follow-ups, which we developed in collaboration with the financial investment research firm Citrini Research. Teams at Adaptyv Bi
Of these, 17 met our criteria. We’re releasing all 211 designs and their data as a resource for the field. In August, we showed how Claude designed protein minibinders after being given a single prompt.
- [2]Claude did not invent scorpion antivenomThe Verge: AI · Oct 22, 12:21 PM
An investment research firm wanted to prove AI could revolutionize drug development. So they teamed up with Anthropic to see whether an unreleased version of Claude could make scorpion antivenom. It couldn't.
It came up with some promising starting points, but nothing close to an actual drug, experts tell The Verge. Citrini Research, the firm behind the experiment, admitted that "none of us could claim to know anything about AI as it related to biology." That didn't stop it from trying to make a drug from scratch - or publishing a blog post titled "We Wanted to Use Claude to Create a New Drug.
- [3]Claude designed protein binders against scorpion neurotoxinsHacker News (AI, 100+ points) · Oct 20, 05:48 PM · same content
AI summary (gemini-3.5-flash-lite); 100% of sentences verified against the sources.
Hype check
Independently evaluated1 independent look since
- Evaluation · Oct 21, 2026VeriEvolve: Evaluator-First Evolution for Tool-Using Agents
- Follow-up · Oct 20, 2026Voxtral 3
- Follow-up · Oct 20, 2026Combining Long-Term Memory and LLMs Comes at a High Cost: From Prompt Injection to Persistent Compromise
- Follow-up · Oct 20, 2026Dell’s security chief warns AI is lowering the barriers to sophisticated cyberattacks
- Follow-up · Oct 20, 2026A Bank Built Its Own Coding Agent. It Took Off — Felipe Felisola Caso & Francesco Galletta, Monzo
- Follow-up · Oct 20, 2026Battle Drones - a vibe coded RTS with audio-based multiplayer
Before this
- Oct 18, 2026Investing in the public benefit
- Oct 16, 2026How employment is changing at Anthropic
- Oct 16, 2026Designing more efficient quantum circuits with Claude
- Oct 16, 2026Claude reveals ECDSA curve origins
- Oct 16, 2026Working at the frontier shapes expectations for the future
- Oct 15, 2026Anthropic gives outside AI watchdog “employee-like” access
