OpenAI Accused Again of Training AI on Private Chats
Your private conversations may be teaching AI — without you knowing.
Deep Dive
The source is a Reddit post stub: a submission by user SirReal14, with a link and comments placeholders. There is no article text, no claims, no companies named, and no accusations about AI training. Anything beyond that isn't in the source.
Key Points
- A researcher claims OpenAI learned from real user conversations, then called the result a breakthrough — the claim is unverified.
- Training means feeding an AI examples; if your chats are included, your private questions become teaching material.
- Most major chatbots let you switch off chat history and training use, but the setting is usually buried in menus.
Why It Matters
Your everyday AI chats may quietly train the product — check your privacy settings today.