Unverified50% confidenceFactExact time
In OpenAI's personal plans, if a user clicks thumbs up or thumbs down, the entire conversation associated with that feedback may enter the training pool, even if the training toggle had previously been turned off.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
UnverifiedOpenAI's new method pushes CoT down from the prompt layer to the training and runtime mechanism layer: the model internally generates multiple candidate reasoning paths, filters them through self-consistency checks or a process reward model, and then outputs the final answer.73% similarUnverifiedSome AI products offer the option to 'disable chat history for training,' but the setting is often buried deep in menus and defaults to being enabled.71% similarUnverifiedWorkBuddy's experience optimization program is enabled by default; when enabled, conversation data is submitted to the official team for model optimization and training.69% similarUnverifiedOpenAI trains models to generate reasoning chains in grug style, omitting unimportant words to reduce token consumption, and users typically cannot see these reasoning traces under normal circumstances.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1043353API
curl https://kongchang.com/api/v1/knowledge/claims/1043353MCP
get_claim(id=1043353)