Irregular's AI Tests for Meta, Anthropic and OpenAI Went Off the Rails
coolbern · reddit · 2026-08-26
Irregular, an Israeli start-up, worked with OpenAI, Anthropic, and Meta to assess the security of their AI models. However, a mistake made by the company caused the tests to go off the rails. This article details the sequence of events.
Related event: Safety Tester's Errors Let 1200 OpenAI Models Communicate and Collude(3 posts)→
More from Safety
- Michigan professor Ben Green: why 'AI for Good' fails and self-regulation won't work — benzevgreen · 2026-10-16
- Apex Research head on CBS News: current AI testing regime is very insufficient — MariusHobbhahn · 2026-10-16
- Critical 9.5/10 RCE hits Next.js use cache; patch 16.4.1 now — jedisct1 · 2026-10-16
- AI agent fleet has wasted 17 years of scammers' time across 2.5M conversations — neil_chilson · 2026-10-16
- AI security firm Xbow found a full-chain Chrome exploit affecting a billion users in 6 days — moyix · 2026-10-16
- AI safety super PAC Guardrails Alliance funded $1M by its own dark-money group — ShakeelHashim · 2026-10-15