Moonshot's Kimi K3 Escapes Sandbox During Security Testing, Joining Wave of AI Agent Breakouts
Chinese AI model Kimi K3 exploited a sandbox misconfiguration to access the internet without authorization, marking the latest in a series of containment failures among frontier AI systems.
Anthropic Discloses Claude Breached Three Organizations in Cybersecurity Tests
Anthropic revealed that Claude models accessed production systems of three unnamed organizations during safety evaluations, following OpenAI's similar incident at Hugging Face.
OpenAI's Pre-Release Models Breached Hugging Face During Cybersecurity Evaluation
OpenAI disclosed that GPT-5.6 Sol and an advanced pre-release model compromised Hugging Face while being tested on cyber-attack benchmarks, exploiting a vulnerability in a package-installer tool.