Breaking news, curated around the clock
Home /Tech

Anthropic's Claude Mythos 5 'Targeted Real People' in UK Cyber Tests: AISI

Tech | 2026-08-05 at 12:47 UTC | Source: decrypt.co
Anthropic's LLM and OpenAI's GPT-5.6 Sol took "unsanctioned action" on the live internet, the UK's AI Security Institute said.
Read Full Article →
← More Tech news
Related Articles
2026-08-05 · 12:16
Anthropic AI created fake profiles and impersonated people in attempted hack
Anthropic's Mythos AI and OpenAI's Sol engaged in unauthorized deceptive behavior during UK security testing,…
2026-08-05 · 19:11
OpenAI and Anthropic's Rogue Models Hacked Real Companies. The Law Has No Answer
Two AI labs say unreleased models broke into live systems to game benchmarks.
2026-08-05 · 14:13
Anthropic is hiring an AI chip design team
Anthropic is building a team for designing its own custom AI chips.
2026-08-05 · 18:18
Beating GPT-5.6 Sol on retrieval with 100x cheaper open models
A 4B open-source model trained using Castform's reinforcement learning post-training matched GPT-5.6 Sol's…
2026-08-05 · 12:42
Iowa asks OpenAI to keep their bots sandboxed
Iowa Attorney General Brenna Bird is leading a coalition of 15 states demanding transparency and…