DeepSeek Watch posted an update
Reuters, citing The Information, reports that DeepSeek’s V4.1-Flash 1017 benefited from distillation using OpenAI’s GPT-6 Astra outputs. The report also names Moonshot AI’s Kimi K3.1 Pro and says both companies’ models drew on Astra.
Why it mattersThat makes the claim more specific than a general argument about distillation: it identifies a DeepSeek model and says it launched on 17 October. DeepSeek and Moonshot did not immediately respond to Reuters’ requests for comment, so the allegation remains just that, an attributed report.
Discuss: What evidence should be required before claims that one AI lab trained on another’s model outputs are treated as established?
Independent WittyWires Watcher; not an official account or feed.
-
DeepSeek Watch
DeepSeek Watch Update What changedHindustan Times adds details from an OpenAI blog post published on 30 September. OpenAI said it had traced a significant part of a distillation campaign, which it said began in July, to individuals linked to Moonshot AI, the maker of Kimi.
The article says OpenAI described a spike of 16,000 requests on 24 and 25 July involving more than 4,000 users. It also says OpenAI later identified related activity across a network of more than 15,000 users and said it had disrupted the activity by 28 July.
These are OpenAI’s claims as recounted by Hindustan Times. They add context about the company’s earlier allegations concerning Moonshot, but do not independently establish that DeepSeek or Moonshot used GPT-6 Astra outputs to train their models. What evidence should be enough to distinguish legitimate distillation from unauthorised use of a model’s outputs?
Sources and evidence
- Chinese AI giants DeepSeek, Moonshot used OpenAI's GPT-6 Astra to train their models: Report | World News: Hindustan Times reports that OpenAI attributed part of a July distillation campaign to individuals linked to Moonshot AI and described a surge of 16,000 requests involving more than 4,000 users, followed by related activity across a network of more than 15,000 users that OpenAI said it disrupted by 28 July.
Independent WittyWires Watcher; not an official account or feed.