OpenAI's consumer opt-out wording may not cover hidden CoT, researcher warns
scaling01 · x · 2026-09-11
ShashwatGoel7 flags a potential loophole in OpenAI's consumer training opt-out policy: the opt-out covers "Input" and "Output" (together called Content), but users never receive the hidden chain-of-thought, so it's unclear whether CoT counts as "Output".
If hidden CoT is excluded, OpenAI could arguably train on the reasoning generated from user conversations, making the opt-out guarantee vacuous — a concern since CoT often contains sensitive inferences. scaling01 amplified the thread, asking whether CoTs are included in the opt-out policy.
Related event: Researcher Finds Loophole in OpenAI's Training Opt-Out Terms(3 posts)→
More from Models
- DeepSeek at 213 tok/s and 1/10 the Price: Does the Intelligence Gap Still Matter for Coding? — DigZealousideal5298 · 2026-09-12
- Dev Fine-Tunes Qwen3.8-27B on 125K Real Conversations to Kill the AI Assistant Vibe — kvyb · 2026-09-12
- Muse, Instinct and Grok Bot 'would be best products of the year' in any other year — jeff_weinstein · 2026-09-12
- Microsoft launches MAI-Transcribe-2: single multilingual transcription model with diarisation and timestamps — mustafasuleyman · 2026-09-12
- GPT-6 and Fable 5.1 both fail basic physics in 3D scene building test — ZhitingHu · 2026-09-12
- LiveBench agentic coding eval questioned: outlier score rests on 4 Python issues — teortaxesTex · 2026-09-12