GPT-6 Astra reads wrong-keyboard-layout input but its safety checks don't catch it; Opus 4.1 too
Sauers_ · x · 2026-09-14
A user found that GPT-6 Astra can infer what you meant when typing with the wrong keyboard layout — with no tools or caveats — yet its safety checks aren't smart enough to detect the same workaround. Opus 4.1 reportedly exhibits the same behavior. The mismatch between model comprehension and guardrail detection remains an open gap for OpenAI and Anthropic to fix.
More from Models
- Rob LeClerc: OLMo 3.1 is the best open model, call for hyperscalers to back a transparent alternative — robleclerc · 2026-09-14
- Analyst: OpenAI's Navier-Stokes model is likely the restarted paused frontier RL run — soumitrashukla9 · 2026-09-14
- Early reviews of Meta's Muse spark prediction it will hit 1B users first — RihardJarc · 2026-09-14
- If MiniMax pulls 70% API margins, Anthropic and OpenAI may be at 90%, argues X thread — teortaxesTex · 2026-09-14
- Aurora1.0, a 150M open model trained on 7B tokens, matches GPT-2-Small — Tall_Abrocoma_3533 · 2026-09-14
- iOS 27 private hooks let apps swap Siri's AI backend with third-party models like Claude — amplifiedamp · 2026-09-14