Xiaomi is a Chinese consumer-electronics manufacturer that also develops AI models under the MiMo name. Its model work is documented here only through the distillation findings below; funding, compute, benchmark performance and the MiMo family's architecture are not established.
MiMo models
MiMo-V2-Pro is the version named in the sources held. Anthropic reported in September 2026 that Xiaomi "may have launched its MiMo-V2-Pro model with a free trial period — which was then extended — with the intent to use the surge in international developer use of the model to distill Claude capabilities," noting that "the bulk of the distillation attacks on Claude began just as the trial period was ending" (Detecting and countering misuse of AI: September 2026 (Anthropic threat intelligence report)). Anthropic states this as an inference from timing rather than as an established intent.
Distillation findings
Anthropic's September 10, 2026 threat intelligence report designates the campaign GTG-16008 and describes it as replaying user conversations and coding sessions from Xiaomi's own MiMo models through Claude, "often run through OpenClaw and OpenCode coding harnesses," to generate data for supervised fine-tuning and reinforcement learning. The report records more than 400,000 requests to Claude across more than 1,500 accounts via proxy services over 20 days in March and April 2026. Claude was used to reconstruct developer environments from exchange transcripts, convert multi-turn conversations into cleaner exchanges, generate both sides of synthetic developer-model dialogues, and judge answer quality (Detecting and countering misuse of AI: September 2026 (Anthropic threat intelligence report)).
Unlike Moonshot AI and DeepSeek, Anthropic's investigation "did not indicate that Xiaomi used Claude's responses to serve its users," but found that it saved exchanges between its own customers and its models and replayed them. The relayed traffic "included sensitive data from users that accessed Xiaomi's models through third-party model routing platforms," containing "names, contact information, corporate data, and other sensitive data from hundreds of Xiaomi users in at least a dozen languages"; Anthropic states it has "no indication US persons' data was exposed," while noting those platforms are commonly used in the United States and Europe. See Adversarial Distillation, AI and Privacy.
Xiaomi is not among the six firms named in the September 8, 2026 NSA/CISA/FBI joint advisory AA26-251A (Joint Cybersecurity Advisory AA26-251A: China-Based AI Companies Conducting Industrial-Scale Distillation Campaigns (NSA/CISA/FBI, September 2026)).
Relationships
- instance-of: Adversarial Distillation — named by Anthropic as one of seven PRC-based labs distilling Claude.
- related: Detecting and countering misuse of AI: September 2026 (Anthropic threat intelligence report) — the report naming the company, its method and its volumes.
- related: DeepSeek, Moonshot AI — the two other labs Anthropic reports as feeding their own users' conversations into Claude.