Only open-weight self-hosted models helped during breaches, Hugging Face and OpenCode found
ccerrato147 · x · 2026-10-23
Continuing the 'Black October' thread: when Hugging Face was breached in July, frontier models' safety filters refused to help analyze the attack. OpenCode got the same refusal mid-breach this month, and only received help after OpenAI admitted it to a vetted defender program. The only help on the spot came from open-weight, self-hosted models. The author's conclusion: a guard that won't fight at night is not a guard — a hosted model answers to its maker, one in your own keep answers to you.
Related event: OpenAI's Security Refusals Fuel Push for Open-Source Models(2 posts)→
More from Models
- 4 Hours of LLM Startup Research Produced a Convincing but Nonsense Product Thesis — annetgriffin · 2026-10-27
- Dev argues Claude burns far more tokens than GPT models for the same coding work — haider1 · 2026-10-27
- MiniMax matches Claude Opus 4.5 in non-benchmaxxed evals, users report — teortaxesTex · 2026-10-27
- Real-world math: GPT-6.1 Terra's half price is offset by 2-3x token usage, costing more than Sol — lordpuddingcup · 2026-10-27
- Fable 5.5 rumored to ship this week as partners still test — kimmonismus · 2026-10-27
- GPT 6.1 Terra vs Claude Sonnet 5.5: speed and value vs raw capability — teortaxesTex · 2026-10-27