News · 22 April 2026

Anthropic's Most Dangerous AI Was Accessed Unauthorised

Anthropic is investigating unauthorised access to Claude Mythos Preview, its AI model deemed too dangerous for public release. The breach occurred through a third-party contractor environment on April 7, the same day Project Glasswing launched. The group guessed the model's location using Anthropic's URL patterns. Anthropic says no core systems were affected, but the security breach raises serious concerns.

By BriefTea Editorial · 22 April 2026 at 12:53

Reported by The Indian Express · How we write briefs

Get the app

More briefs

WorldTurkey, Saudi Arabia and Pakistan to discuss new mutual protection deal on Monday WorldIran's unseen supreme leader urges Gulf countries to fight a shared enemy BusinessIceland supermarket boss tells the Chancellor to cut business taxes for employers GeneralNigel Farage's £5m donor has major stakes in arms industry PoliticsTreasury officials advise John Healey to tax banks and oil companies more HealthCorrecting hearing loss can cut dementia risk by almost half Browse all stories →