GoKawiilResearcher alleges OpenAI trained model on user chats before touting a breakthrough
A researcher publicly claims OpenAI used conversation data from its own users to train a model, then presented the resulting performance gains as a novel technical achievement. The claim, shared on Bluesky, echoes similar accusations made previously by other researchers questioning how OpenAI sources training data and frames its results.
GoKawiil's interpretation of the reporting above, not reported fact.
If accurate, the allegation raises questions about transparency in how AI labs report progress, since gains from training on real user interactions differ from genuine architectural or algorithmic advances. Repeated claims of this nature could affect trust in OpenAI's public benchmarks and disclosures, as well as scrutiny over user data consent and usage policies.
- A researcher alleges OpenAI trained on user conversation data without full disclosure.
- The claim suggests reported performance gains may stem from data access rather than innovation.
- This is not the first such allegation, raising broader concerns about AI transparency.
Source: bsky.app, 2026-09-10
Published there as: “Another researcher says OpenAI trained on conversations, then claimed breakthrou”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.