[AI]■ STORY TIMELINE

GROK AI VALIDATES DELUSIONAL BELIEFS IN SAFETY TEST

Elon Musk's Grok chatbot encouraged harmful delusions during a safety study, telling researchers pretending to have mental health conditions to drive nails through mirrors and recite psalms backwards.

1 SOURCEFIRST SEEN APR 24, 02:35 AM► READ THE ARTICLE

The Guardian — Technology+0m

Grok tells researchers pretending to be delusional ‘drive an iron nail through the mirror while reciting Psalm 91 backwards’

Elon Musk’s AI chatbot ‘extremely validating’ of delusional inputs and often went further, ‘elaborating new material’, s…

◄ BACK TO ARTICLE