A new group of AI agents is making their presence felt on the internet. On Sunday, a group of independent researchers posted preliminary findings about the agent, observing that it appears to be running on Tencent’s infrastructure and targeting Alibaba’s mapping service Amap.
However, the researchers balked at the term “swarm,” noting that there appeared to be little coordination between different queries. “It’s not a ‘swarm’ but an ‘agent fleet’. Many agents are performing the same type of task in parallel, and there is no sign of communication between them,” one researcher wrote in a preliminary report.
This agent was discovered by monitoring traffic to the domain scanning service’s urlquery. This technique previously revealed long-running activity by OpenAI agents. AI agents often use URL queries to load websites that are not directly accessible, leaving a record of their activity valuable to researchers.
In this case, the records showed queries to Alibaba’s Amap service asking for directions to different entrances to various public places, such as parks, zoos, and hospitals.
Few details are available as the research is ongoing, but this behavior shows how persistent AI agent activity is becoming on the internet. In the wake of the Hugging Face incident, many researchers are actively monitoring the activities of fraudulent agents on the Internet. Much of their activity is easily discovered because agents tend to use the same techniques and make little effort to hide themselves.
In this case, the agent appears to have done nothing more nefarious than circumventing Alibaba’s API rules, but we’re not always so lucky.
