# Disrupting a coordinated model-distillation campaign

OpenAI attributed a core cluster of the extraction attempts involving over 15,000 accounts to individuals associated with Moonshot AI, the developer of Kimi.

- Operators manipulated model interactions and replayed encrypted reasoning across conversations to coerce models into decrypting hidden reasoning.
- The campaign peaked on July 24 and 25 with 16,000 attempted extraction requests from 4,000 users before full disruption on July 28.
- OpenAI deployed mitigations including banning accounts, closing reasoning replay pathways, and adding checks to hold streamed output exposing reasoning.
- The company shared threat findings with industry partners through the Frontier Model Forum and government channels.

## Why it matters

AI developers face risks of model capability theft and safeguard bypasses through adversarial distillation, requiring defenses across reasoning replay artifacts and streaming outputs.

## Sources

- [OpenAI: Disrupting a coordinated model-distillation campaign](https://openai.com/index/disrupting-a-coordinated-model-distillation-campaign)

---

Summarized by dstilled on 2026-09-30. https://dstilled.ai/story/d8f65008-da1d-4e3e-991a-3a37444bd6be