According to Decrypt, OpenAI stated it disrupted a campaign involving more than 15,000 users and identified a core cluster tied to individuals associated with Moonshot, the startup behind Kimi. The company said those linked to China’s Moonshot initiative attempted to copy its AI’s hidden reasoning processes. No further details on methods, timeline, or direct evidence were provided in the report.
OpenAI described the effort as targeting replication of internal reasoning mechanisms not publicly disclosed. The affected group was linked to people connected to Moonshot, which develops the Kimi AI. The disruption occurred after OpenAI identified indicators of unauthorized access attempts or probing behavior. The company did not specify whether legal action was pursued or if technical defenses were enhanced in response.
The announcement reflects OpenAI’s ongoing vigilance regarding external attempts to reverse-engineer proprietary AI functionality. Officials emphasized compliance with security protocols and cooperation with relevant authorities where applicable. No names of individuals or specific locations were released.
