The archive provides Chinese and English research texts. Translations may be AI-assisted and may not yet have received independent human review. Consult the original-language sources alongside them.
Anthropic
Read original source (EN)https://www.anthropic.com/threat-intelligence-report-september-2026
Reading location
Lines 1397–1510: distillation chapter; date separately checked against the official index.
Only the distillation chapter and introduction were checked. Seven-laboratory attribution, scale, user data and bans are company allegations or accounts. They were not independently verified, and named parties' responses were not obtained.
Capture status
No verified capture is recorded. The original page may change; an access date is not a historical version date.
Events citing this source
Anthropic report expands distillation allegations to seven laboratories
Anthropic extended attribution to Alibaba, Moonshot, DeepSeek, Z.ai, Xiaomi, SenseTime and MiniMax, raising issues of resale, forwarding and training controls over user interactions.
Cite this record
ANTI-AI ARCHIVE. “Detecting and countering misuse of AI: September 2026.” src-0244. Include this page’s URL and your access date when citing.