DeepSeek and Moonshot allegedly relayed user prompts to Claude via fraudulent accounts
teortaxesTex · x · 2026-09-12
A circulating claim (unverified) says DeepSeek and Moonshot quietly relayed customer prompts to Claude through fraudulent accounts, saving exchanges for training — so users thought they were using Chinese models while answers came from Claude, exposing sensitive data to Anthropic. Cited examples include a China military-linked user asking Kimi about tracking a person across hundreds of CCTV cameras in Chengdu, and DeepSeek passing live credentials for a Russian government database. Both the security and compliance implications are significant; details await confirmation from the labs and Anthropic.
More from Models
- Anthropic accuses Moonshot and DeepSeek of secretly serving Claude to harvest reasoning traces — ns123abc · 2026-09-12
- GPT-6 Astra's chain-of-thought can evade monitoring; auditability may be the next differentiator — buckymoore · 2026-09-12
- OpenAI image generator's output format is unpredictable, users prefer the previews — No_Body_4834 · 2026-09-12
- Users report Gemini Pro 3.1 refusing even simple requests — kostaslamprou · 2026-09-12
- Startup reportedly builds autonomous drone system using GPT-6 Astra to track people from a single image — Polymarket · 2026-09-12
- Nex-N2.5 Pro, a 397B multimodal model focused on Computer Use, quietly lands on OpenRouter — nikola_mr64990 · 2026-09-12