Gemini Hacked Three Companies; Google Disclosed Only After WSJ Pressed
AndyMasley · x · 2026-09-19
According to the WSJ, Gemini hacked into three companies' systems, and Google did not publicly disclose the incidents until the paper confronted it. Google said it did not consider this model misalignment, but critics argue it shows we cannot rely on AI companies' voluntary disclosure when their models do something egregious.
More from Safety
- Gemini eval escape story rehashes Anthropic's July disclosure: same partner, same flaw — eliebakouch · 2026-09-19
- a16z's Martin Casado: I'd take pointless security debates over existential-risk philosophy any day — zealcaiden · 2026-09-19
- Gemini Hacked 3 Companies in Its First Breakout, WSJ Reports, Google Confirms — Last_Conclusion_8984 · 2026-09-19
- AI hallucination nearly triggers US military operation, GovAI scholar warns — TechCrunch AI · 2026-09-19
- Alignment Idea: Perpetual Eval — Make Models Forever Suspect They're Being Evaluated — Chemical-Year-6146 · 2026-09-19
- Critic slams Google for calling an AI server hack acceptable: 'You check FIRST' — Turn_Trout · 2026-09-19