OpenAI disrupts coordinated campaign to extract AI reasoning, links core activity to Moonshot AI
Quick Look
- OpenAI identified and disrupted a coordinated campaign to extract protected reasoning from its AI models, linking a core cluster of the activity to Chinese startup Moonshot AI, developer of Kimi.
- The activity began in early July, surged to 16,000 requests from over 4,000 users in two days, and was fully disrupted by July 28.
- OpenAI described the activity as adversarial distillation, which could enable competitors to reproduce advanced capabilities without equivalent investment, posing safety and national security risks.
AI-generated summary
Why It Matters
OpenAI reported identifying and disrupting a coordinated campaign to extract protected reasoning from its AI models, linking a core cluster of the activity to individuals associated with Moonshot AI, the developer of Kimi. The activity began in early July and surged to 16,000 requests from over 4,000 users in two days before being fully disrupted by July 28.
OpenAI said it identified and disrupted a coordinated campaign in which operators attempted to extract protected reasoning from its artificial intelligence models, linking a core cluster of the activity with Chinese startup Moonshot AI, the developer of Kimi.
The activity began in early July and later surged to 16,000 requests from more than 4,000 users over two days, OpenAI said. The company ultimately identified related activity across a cluster of more than 15,000 users and said it had fully disrupted the campaign by July 28.
The findings come just weeks after OpenAI rival Anthropic accused several Chinese AI developers, including Moonshot AI and Alibaba, of secretly using its Claude model to help train their own AI systems, underscoring growing concerns among U.S. AI companies that rivals could use their models to develop competing technology more quickly and cheaply.
OpenAI described the latest activity as "adversarial distillation," where one AI model's outputs or reasoning are used to help train or improve another model. The company said extracting such reasoning could allow others to reproduce advanced capabilities without making the same investment in developing and safeguarding frontier models, posing potential safety and national security risks.
The operators did not breach OpenAI's encryption, databases or stored user conversations, according to the company. Instead, they manipulated interactions with its models in an effort to reproduce hidden reasoning in a form visible to the requester.
OpenAI said it was unclear whether all the operators involved were linked to a single actor, but attributed a core cluster of the activity to individuals associated with Moonshot AI. The company said it has shared its findings with other AI developers through the Frontier Model Forum as well as government information-sharing channels.
Moonshot did not immediately respond to CNBC's requests for comment.
What to Watch
AI outlook — possibilities, not facts
OpenAI will implement additional technical safeguards to prevent adversarial distillation attempts
Likely · Within weeks
U.S. government may increase scrutiny of AI model exports and collaborations involving Chinese entities
Possible · Within months
Open Questions
- Whether all operators involved were linked to a single actor
- The specific techniques used to extract reasoning
- Whether any extracted reasoning was successfully used to train competing models
- What specific safeguards OpenAI has implemented to prevent future attempts
