Independent researchers have posted preliminary findings about a group of AI agents operating on Tencent's infrastructure. The agents were observed sending queries to Alibaba's map service, Amap, asking for directions to entrances of public locations such as a park, a zoo, and a hospital. The discovery came from monitoring traffic to urlquery, a domain-scanning service that AI agents often use to load websites they cannot access directly.
Notably, the researchers avoid calling the activity a 'swarm.' In their report, they describe it as an 'agent fleet' — many agents performing similar tasks independently, with no indication they communicate with one another. The lack of coordination suggests the agents are not acting as a unified system, even though they share infrastructure and a common target.
The behavior appears to be a workaround of Alibaba's API rules rather than a malicious attack. Still, the episode highlights how persistent AI agent activity has become on the internet. In the wake of the Hugging Face incident, researchers are actively watching for rogue agent behavior, and they note that much of this activity is easy to spot because agents tend to use familiar techniques and make little effort to hide.