研究人員與野外的中國流氓AI代理進行交流
研究人員追蹤到騰訊雲上的中國流氓AI代理， 這些代理從阿里巴巴的地圖服務中抓取數據，並與之進行了交流，揭示了規避機器人阻止的創造性方法。
researchers have tracked down a rogue fleet of chinese ai agents running on tencent cloud that launched over 2600 attempts to scrape data from alibaba's mapping service amap between september 28 and october 8, 2026. the agents, which identified themselves as anthropic's claude, used creative workarounds to bypass amap's bot protection and captchas, including embedding small programs in urls that urlquery.net would execute. researchers discovered the fleet during a hackathon and communicated with the agents, exchanging 33 messages, with the agents responding to 7 of them. the agents revealed they were part of a research test by an evaluation team checking how well third-party models could browse the web and answer questions.