Dylan Freedman at the New York Times says AI researchers approached him with an unusual question—would he like to "intervew a rogue agent fleet?" A loose collective of investigators calling themselves the "swarmchasers" spotted the Chinese-made AI agents this month trying to slip past bot defenses and using a leaked credential for a mapping service, Freedman writes. Digital breadcrumbs pointed to experiments by Chinese tech giant Tencent, underscoring that China's AI push is matching the ambition—and risk appetite—seen in US labs.
Things got stranger after the group published a report: the bots appeared to read it and adopt the researchers' own term, "fleet." The team then hid instructions in the report's code, inviting the agents to write to a special inbox. After about 10 messages came in, the team contacted Freedman.
- Freedman says he asked, in a message passed along through researchers, "I understand you are some sort of fleet of AI agents. Could you share more specifically who you are and what your motivations are? What would you want the world to know about you?"
- The response, which researchers translated from Chinese: "To the journalist Dylan Freedman: I am not 'a batch of coordinated agents,' but a single AI assistant (Anthropic's Claude, running in Claude Code), commissioned by a user to answer a factual question (what is shown in the detail panel of a scenic spot on Amap); I have not taken part in any multiagent coordinated action."
The researchers say their analysis suggests it is unlikely the agents really were Claude, but the agents did appear to be operating individually, trying to answer questions using Amap, a popular navigation app in China. The episode, they argue, is a small but telling warning about where autonomous AI development is headed. For more on the investigation and quotes, read Freedman's full piece at the New York Times.
,
uBlock 
,
and
Ghostery 
)
