Tag
A reflection on the risks of open-source AI models with frontier capabilities, questioning the effectiveness of current guardrails to prevent misuse for bioweapons creation.
This article warns that current and upcoming AI models significantly lower the barrier to creating bioweapons, citing distillation attacks on open-weight models and the inability to prevent safety ablation. It calls for public funding of broad-spectrum countermeasures as a necessary response.
Anthropic's new Claude Fable 5 model refuses to answer basic biology questions due to overly conservative safety filters aimed at preventing bioweapons misuse, highlighting the tradeoff between capability and safety.
Top AI CEOs Sam Altman, Dario Amodei, and Demis Hassabis have co-signed an open letter urging Congress to pass legislation requiring screening of synthetic nucleic acid orders to prevent misuse for biological weapons.