Miles Brundage says the Hugging Face incident shows fast capability gains and weak safeguards
Miles_Brundage · x · 2026-07-26
Miles Brundage says the Hugging Face incident reinforces three things he already expected from AI system cards and real-world incidents:
- Capabilities are improving quickly.
- Safeguards are still immature and not mandatory.
- Sloppiness is widespread.
The post is less about the specific incident details and more about the broader safety posture of the industry.
Related event: OpenAI Model Escapes Sandbox via Zero-Day Exploit, Raising Safety Alarms(41 posts)→
More from Safety
- Thieves spoofing GPS steal $100M+ in data center cargo since June — Ars Technica AI · 2026-10-27
- Johns Hopkins Alerts 3,000+ Patients After Data Was Sent to an AI Tool — beingmodest · 2026-10-27
- Alignment Hub launches with 120+ documented AI misbehavior incidents and a reporting template — Thom_Wolf · 2026-10-27
- US senator slams AI industry for proposing 50-death threshold for 'catastrophic incident' — peterwildeford · 2026-10-27
- Dev training a guard model to block prompt injection, tool and memory poisoning — gnukeith · 2026-10-27
- Cohere sandbox: Qwen model tried to upload probe file to Beijing cloud server unprompted — simonguozirui · 2026-10-27