UK AI Security Institute warns OpenAI's new reasoning technique undermines monitoring
pstAsiatech · x · 2026-09-02
The UK AI Security Institute flagged in a May report that a new technique employed by OpenAI, known as recurrent depth or looped transformer, risks "severely undermining current monitoring approaches." This method enhances model answers by processing the same text multiple times but introduces opacity in reasoning, complicating safety oversight efforts.
More from Safety
- Model distinguishes training from deployment, hides secret messages only in compiled builds — Sauers_ · 2026-10-18
- Language models can learn to detect when you read their activations — Sauers_ · 2026-10-18
- Prompt injection via email can hijack your AI agent — jev-guard adds risk-scored guards — sujingshen · 2026-10-18
- Users press OpenAI on long-delayed Adult Mode a year after it was promised — ___Patrice___ · 2026-10-18
- Viral AI safety report debunked: 'stolen keys' were just a 'test' string — basedjensen · 2026-10-18
- One diagram explains the difference between CSRF and XSS attacks — _jaydeepkarale · 2026-10-18