OpenAI says it disrupted a month-long campaign to extract reasoning from its models, with the core activity linked to people associated with China’s Moonshot AI, developer of Kimi. It argues that large-scale distillation could let competitors reproduce advanced capabilities while bypassing the safety safeguards of the original models.

The campaign began on 1 July and peaked on 24 and 25 July, when OpenAI recorded 16,000 requests using an extraction pattern across more than 4,000 users; it found related activity across over 15,000 users. OpenAI says it stopped the campaign on 28 July, banned accounts and tightened controls, while noting it could not establish whether every operator was linked to one company. Moonshot AI had not immediately responded to a request for comment.