Summary
According to a post on X by @seanhn, users shared screenshots claiming that from Nov 29, Daybreak Blue sessions on personal accounts may be used for training even when the training setting is off. The post says this could put unreported bugs and proofs of concept into training data. @moyix says he confirmed it, but OpenAI's notice was not seen directly. The same report notes that OpenAI tightened ChatGPT for teens and will let it end chats with users fixated on violence.
AI-assisted summary based on the listed source.
What happened
Users shared screenshots saying that from Nov 29, Daybreak Blue sessions on personal accounts may be used for training even with the training setting off, which would put unreported bugs and proofs of concept into training data; @moyix says he confirmed it, but OpenAI's notice wasn't seen directly. OpenAI also...
Why it matters
Users of personal accounts who rely on the training setting to limit data use would be directly affected if the claim holds. Security researchers who share unreported bugs or proofs of concept in these sessions could expose sensitive work. The key thing to watch is whether OpenAI publishes an official notice confirming or rejecting the change.
Signal Intelligence
Signal Strength 90%
Technical label WATCH
Public Interest 46
Category BIG MOVE
Reader Depth TECHNICAL
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
Public Interest components
Recognizable Entity Score 82
Practical Impact Score 0
Novelty Interest Score 70
Consequence Score 34
Curiosity Score 0
Shareability Score 59
Why this is here
VQV surfaced this signal because it is recent, relevant to Consumer AI, connected to Scobleizer News Pulse (X).
