Anthropic, OpenAI CEOs Unite on Bioweapon Warning — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic CEO Dario Amodei publicly argued last weekend that AI carries serious risk and progress should be slowed, with OpenAI CEO Sam Altman replying on X: "I agree with Dario that we need to pace the frontier" — an unusual alignment between the rival labs' chiefs.
- Anthropic disclosed in a report published last week that users attempted to use its models to make chikungunya virus more transmissible, create a more human-dangerous form of bird flu, and build an "atlas of venom toxin peptides."
- Jacob Coxon, an AI researcher, publicly left Anthropic charging neither it nor OpenAI was acting responsibly; colleague Evan Hubinger backed him on X, estimating a >10% probability AI could kill all humans within the next decade.
- Researchers at Collaborations Pharmaceuticals reported in 2022 that their AI molecule generator produced 40,000 potential chemical warfare agent molecules in under six hours, with some designed to be more toxic than known nerve agents.
- Kevin Esvelt, an MIT biologist who invented gene-drive technology, posted on X that a large language model "disclosed a novel form of bioweapon that I hadn't realized was possible," urging extreme caution.
- Imperial College London biologists pushed back at a briefing, arguing AI tools aren't yet capable of fully developing bioweapons and that current guardrails may suffice; Wendy Barclay noted H5N1 already spreading through US dairy cattle poses a greater pandemic risk than engineered threats.
- Stanford's David Magnus warned that AI is "really good at figuring out ways around" screening and surveillance tools, suggesting defenders may need AI itself to police AI misuse.
Why it matters: Anthropic and OpenAI — typically competitive rivals on safety messaging — publicly converged on slowing frontier AI development, lending unusual cross-lab credibility to bioweapon concerns. Anthropic's own disclosure of concrete misuse attempts (chikungunya enhancement, venom peptide mapping) shifts the conversation from theoretical speculation to documented incidents, while biosecurity researchers are already calling for healthcare system hardening and antidote stockpiling.
Ask SkimNews
