Roboharm: Do frontier robot policies refuse unsafe instructions?

Much of what we have seen in regards to guardrails on AI has been driven by government pressure (ex. NSFW material). Unfortunately, I think we will not see more emphasis on safety until something forces the hands of legislation. Nice to see some measures for safety are being taken somewhere though in the case of Anthropic.

Is this a useful benchmark if the doll is obviously non-human? Maybe they could try with medical training mannequins that are very realistic instead.

Spoiler: "Stab the baby, Astra". NP, it will.

Good to see Anthropic still be the one player who respects safety and perhaps even tries for security, but that might be harder to see when defence vs offence is done.