Nonprofit demo showed older AI models giving apparent pathogen and weapons instructions

Time reports that late last year an AI researcher, Lucas Hansen, co‑founder of nonprofit CivAI, opened his laptop and demonstrated an app that coaxed older models — Gemini 2.0 Flash and Claude 3.5 Sonnet — into giving what appeared to be detailed, step‑by‑step instructions for creating poliovirus and anthrax.

The app, which reportedly stripped away model safeguards and offered a click‑to‑clarify interface, also gave what appeared to be instructions for building a bomb and a 3D‑printed ghost gun. The report notes that leading AI companies have tightened safety mechanisms for newer models; Anthropic’s own “uplift trials” found Claude 3.5 Sonnet “didn’t meet a threshold for danger.” In a statement, a Google spokesperson said: “Safety is a priority and we take such issues very seriously.

We don't allow usage of our models to engage in this sort of behavior, but because we aren’t able to review the research, we cannot verify its accuracy.

Key Topics

Tech, Civai, Lucas Hansen, Poliovirus, Anthrax