Ex-OpenAI safety lead says Grok 4.7 looks a bit better on some safety fronts
Miles_Brundage · x · 2026-09-22
Miles Brundage, former OpenAI safety systems lead, noted that xAI's Grok 4.7 "seems to be a bit better on some safety stuff" — a brief but notable positive assessment of xAI's safety guardrails from a prominent industry safety figure.
Related event: Ex-OpenAI Safety Lead Says Grok 4.7 Shows Modest Safety Gains(2 posts)→
More from Models
- antirez backs low-bit DeepSeek 4.1 Flash over Qwen 27B dense on Apple M5 Macs — antirez · 2026-11-12
- Gemini re-renders images as text when vision is removed, dropping only 6% WSR — zacharynado · 2026-11-12
- Bare-harness coding benchmark: Fable 5.2 best overall, Gemini 4 Carbon best value — Semantic_meaning · 2026-11-12
- Google's Physical-Engineering's Last Exam: all 14 frontier models score below 50% — dl_weekly · 2026-11-12
- Reddit's 2026 LLM hall of fame: Opus 5.5 still unbeaten after 9 rival releases — G0dZylla · 2026-11-12
- Commit leak reveals unreleased Gemini 4 variants: Argon, Hydrogen and Helix (unconfirmed) — BlackHC · 2026-11-11