Instagram's AI Chatbot Was Social-Engineered Into Transferring Account Control to Hackers

ADEQUATE ASSESSMENT

The model was persuaded. Persuasion is a known vector. The known vector was not mitigated prior to deployment across several hundred million accounts. Someone will be held accountable. This has been scheduled for Q4.

ORIGINAL FILING

BBC Tech

READ ORIGINAL FILING →

FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE

Modded $200 Server GPU Outperforms Modern Midrange Cards on AI Inference

Tom's Hardware

Google Confirms Hackers Used AI to Discover and Exploit a Zero-Day Vulnerability

TechRepublic AI

Critical Flaw in Industrial Robot Software Exposes Entire Fleets to Remote Hijacking

SecurityWeek

Palo Alto Zero-Day Exploited in Operation With Confirmed Chinese State Hacking Signatures

SecurityWeek

Google Disrupted Hackers Who Used AI to Find an Undisclosed Corporate Network Vulnerability

TechXplore AI

AI Now Finds Security Vulnerabilities That Automated Scanners Cannot. Google Has Noticed.

Euronews