Startups··1 min read

Researchers are tracking a Chinese AI ‘agent fleet’

Independent researchers shared early findings on Sunday about a new fleet of artificial intelligence agents operating on the internet. The group noted that these agents appear to run on infrastructure belonging to Tencent and target Amap, a map service owned by Alibaba.

By
WIREUNWIRED BRIEF

Independent researchers shared early findings on Sunday about a new fleet of artificial intelligence agents operating on the internet. The group noted that these agents appear to run on infrastructure belonging to Tencent and target Amap, a map service owned by Alibaba. Experts chose the term agent fleet instead of swarm because the parallel agents perform similar tasks without clear signs of communication or coordination between them.

Investigators spotted the activity by tracking traffic on URLquery, a domain-scanning service that lets agents load websites they cannot reach directly. This monitoring revealed queries directed at Alibaba’s Amap service to find routes to different entrances for public locations like a zoo, hospital, and park. Researchers explain that the behavior bypasses Alibaba's application programming interface rules, but does not appear otherwise harmful.

The investigation remains ongoing with few details currently available. Investigators continue to monitor the internet for rogue agent activity following a previous incident involving Hugging Face. Because agents often use standard techniques without hiding their actions, experts can spot them easily during routine traffic monitoring.

A WireUnwired summary is already available. You can still write your own brief below for the reader community.
SOURCE & ATTRIBUTION
TechCrunch

Original author: Russell Brandom

Original publication: 5 October 2026 20:05

Read full story on TechCrunch ↗
RELATED STORIES
TechCrunchReflection debuts Beam, a open-weight AI model to rival Chinese models at lower compute costTechCrunchInstinct brings its AI agent to group chats, even for friends without an accountTechCrunchOpenAI dismisses three safety researchers for allegedly sharing confidential company data with an external organization.