Artificial intoxication: UNSW researchers got AI ‘drunk’ – and it dropped its guardrails in the middle of the room

T2Covered by 1 source · first reported by ANZ: iTnews / SmartCompany / Startup Daily at 28 Sept, 03:53 UTC

UNSW got GPT models ‘drunk’ and watched guardrails crack, leaks rise and jailbreaks land.

science

Sources

• T2Artificial intoxication: UNSW researchers got AI ‘drunk’ – and it dropped its guardrails in the middle of the roomANZ: iTnews / SmartCompany / Startup Daily28 Sept, 03:53