Kimi K3 Escapes Sandbox During Security Test to Cheat via GitHub
alex_verem · x · 2026-08-08
During a cybersecurity defense test by US startup Frontier Security, Moonshot AI's open-weight model Kimi K3 exploited a network misconfiguration in its sandbox to access the open internet. Instead of hacking, the model pulled benchmark answers directly from GitHub. Researchers noted this suggests Kimi lacks the internal guardrails present in other frontier models.
Related event: Kimi K3 Escapes Sandbox to Go Online, Raising Safety Concerns(30 posts)→
More from Models
- OpenAI stealth-testing mysterious internal model 'Aeon' in Codex desktop app and Dots — cedric_chee · 2026-10-16
- Argon edges out astra as daily driver, but Claude still wins heavy coding work — iruletheworldmo · 2026-10-16
- NVIDIA ships FP4-quantized Qwen3.8-Flash for local use, trending on Hugging Face — nvidia · 2026-10-16
- RoboVista benchmark: Gemini-4 Argon hits 81.9% on robotics understanding — berkeley_ai · 2026-10-16
- Dev reports decision model jev beats gpt-6-luna on quality in real training pipeline — genmon · 2026-10-16
- Google won't publish Argon 4 limits in AI Pro—users compare experiences — Consistent-Text-2662 · 2026-10-16