ReferIndia News Anthropic Is Working On A 'Switch Button' To Make AI Forget Its Most Dangerous Skills

ReferIndia News

ePrescribe

Clinic chalana ab hoga super easy—smart software ke saath!

Contact Now
News Image

Anthropic Is Working On A 'Switch Button' To Make AI Forget Its Most Dangerous Skills

Published on: July 9, 2026, 3:36 p.m. | Source: Times Now

Anthropic has unveiled GRAM, an experimental AI safety system that could let models "forget" dangerous knowledge like hacking and virology while keeping everyday capabilities intact., Technology & Science, Times Now

Checkout more news
Ad Banner

✅ Secure, cloud-based clinic management software

Manage digital prescriptions, appointments, patient records, billing, files, and reports — everything your clinic needs in one powerful platform.

Know more
ReferIndia News contact