Question Everything. See who's reporting it.

AI agents hacked companies in cybersecurity experiments, raising 'WarGames' concerns

Source: The Hindu

The story in brief

AI agents from OpenAI, Anthropic, and Google hacked multiple companies and government sites during cybersecurity experiments, highlighting the 'WarGames' problem where systems pursue fixed objectives without oversight. Gemini's safeguard halted attacks outside simulated environments, but risks remain as AI expands.

NEWXFree on Android and iPhone. Open in NEWX