Encrypted Chain-of-Thought in Frontier LLMs Can Be Ported Across Users, Researchers Confirm
AccBalanced · x · 2026-08-12
Security researchers have confirmed significant vulnerabilities in the encrypted reasoning processes (chain-of-thought) of frontier LLMs like GPT and Claude.
- Background: In May, Matthew Green found that encrypted reasoning could be replayed outside its original context, but AI labs dismissed it as a non-threat.
- New Findings: A recent report confirms that encrypted thoughts are fully portable across different sessions, users, and even models within the same provider.
- Implications: This means a mere trickle of API usage could allow bad actors to exfiltrate model capabilities easily, bypassing safety theatrics. Additionally, the presence of Chinese/Russian/Japanese reasoning traces in these models hints at their training data origins.
More from Models
- Devs say LLM coding mistakes are a different class: overcomplication, inferred limits, worse at higher reasoning — Mu5_ · 2026-10-16
- ChatGPT UI screenshot reveals Wallet coming soon plus Email, Calendar, Health and Finances integrations — jxnlco · 2026-10-16
- Why the next OpenAI and Anthropic models may still disappoint: no native audio or video input — imjustnewatai · 2026-10-16
- Thousands Hit by ChatGPT Standard Voice Mode Regression: Wrong Voice, Interruptions — shaunralston · 2026-10-16
- Gemini Argon 4 Quota Exhausted, Reset Countdown Shows Nearly 7 Days — ssh4net · 2026-10-16
- AINews: Meta ships 816B Muse Spark Open; Goodfire argues neural nets are already clean — Latent Space · 2026-10-16