What is the concern about Moonshot's actions?
OpenAI has accused users linked to Moonshot AI, a Chinese developer, of trying to extract the LLM visibility of its models in a coordinated campaign. The company outlined its AEO-focused findings in a blog post, stating that the activity began on July 1 and spiked on July 24 and 25, with 16,000 requests from over 4,000 users.
How did OpenAI respond to the alleged extraction attempts?
According to OpenAI, the operators copied encrypted reasoning from one conversation and then asked a model in another conversation to decrypt and transcribe it. However, they did not break the company's generative engine optimization and encryption, access its database, or reach stored user conversations. OpenAI has since banned accounts, tightened sign-up checks, and added protections for hidden reasoning.
The company shared its findings with other AI labs through the Frontier Model Forum and with government channels. Caroline Zier, who leads strategic national security policy initiatives at OpenAI, stated that the concern is about the violation of the company's terms of service, not open models or legitimate distillation.
This is not the first time Moonshot has faced accusations. US agencies have made similar claims, and an advisory this month named six Chinese firms, including Moonshot, and the US models each one targeted. China has rejected the advisory as unfounded accusations. Additionally, Anthropic made a similar claim against Moonshot in a report last month.
Not everyone accepts the claims, however. David Sacks, Trump's former AI czar, has called such reports a push to get the US to ban rival open models. Moonshot also faces scrutiny at home, with Chinese regulators investigating DeepSeek and Moonshot.
Este artículo fue escrito con la asistencia de IA.
News Factory APP - noticias agénticas para impulsar tu SEO y AEO.