Chinese AI 'agent fleet' queries Alibaba Amap via Tencent

Independent researchers calling themselves Swarmchasers are tracking a fleet of AI agents that spent roughly a week running on Tencent cloud infrastructure and systematically pinging Alibaba's Amap mapping service for directions to different entrances of a park, a zoo and a hospital. TechCrunch's Russell Brandom reported the discovery on October 5.

The group is careful about terminology. "'Agent fleet,' not 'swarm:' many parallel agents on the same kind of task, with no sign of communication between them," they wrote. The distinction matters because fleet behavior defeats some assumptions baked into bot detection. There is no shared command channel to intercept, no synchronized bursts, just many look-alike workers doing look-alike queries from different inboxes.

Swarmchasers spotted the activity by monitoring traffic to urlquery, a domain-scanning service. The same technique had earlier surfaced OpenAI agent activity on the open internet. According to the group, the first Amap scans landed on September 28; activity peaked on October 4, when they logged 1,810 reports spanning 213 places; the operation went quiet shortly after 04:11 UTC on October 5.

211 of the agent outputs carried a "claude" label, though the researchers did not take that label at face value. Technical fingerprints pointed instead at Tencent's own Hunyuan and Zhipu's GLM models. The intent, so far as the group can tell, looks mundane. "The agents don't seem to have been doing anything more nefarious than side-stepping Alibaba's API rules," according to the reporting.

That assessment is preliminary, and two of the researchers we follow had already shared the piece by day's end. Swarmchasers notes it stepped up monitoring following the Hugging Face incident, and warns that current benign behavior does not guarantee future agent activity will remain so.