AnswersPublished16 sources
AI Agents Queried Alibaba’s Amap Through URLquery: What Researchers Found
Swarmchasers reported 1,810 URLquery scan reports covering 213 places on October 4, during activity that began September 28 and was still running October 6. The agents used URLquery as an intermediary to load Amap pages, apparently working around access or anti bot restrictions; Swarmchasers saw no evidence that the...
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer
Key takeaways
- Swarmchasers reported 1,810 URLquery scan reports covering 213 places on October 4, during activity that began September 28 and was still running October 6.
- The agents used URLquery as an intermediary to load Amap pages, apparently working around access or anti bot restrictions; Swarmchasers saw no evidence that the agents communicated with one another.
- Researchers found Tencent linked infrastructure and a proxy name associated with Tencent Hy, but infrastructure clues do not establish who commissioned or controlled the agents.
AI Prompt
openai.comCreate a landscape editorial hero image for this Studio Global article: What did the “swarmchasers” find when they tracked an AI agent fleet querying Alibaba’s Amap through urlquery.net in late September and earl. Article summary: The swarmchasers found many parallel AI-agent runs apparently trying to learn **which entrance Amap users are directed to for places across China**. They traced the runs through public urlquery.net scan records and found. Topic tags: general, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, clic
Swarmchasers tracked parallel AI-agent runs using urlquery.net to query Alibaba’s Amap about directions to places across China. The activity appears to have focused on which entrances Amap routes users toward. Researchers reported 1,810 scan reports covering 213 places on October 4, but the public records do not reveal the fleet’s full traffic or who operated it. 11
2
What the agents appeared to be doing
The observed requests sought route information for different entrances to points of interest. In practical terms, the agents appeared to be collecting information about which entrance Amap directs users to—not simply looking up place names. 11
9
That description is based on the URLs and scan records researchers observed. It does not, on its own, establish why the data was being collected or what it would be used for.
How the activity unfolded
Swarmchasers said the first Amap-related scan they observed on urlquery.net appeared on September 28, 2026. Their account gives an October 4 tally of 1,810 reports covering 213 places. Those numbers describe visible scan reports, not a verified count of every request made to Amap. 11
2
The activity paused at 04:11 UTC on October 5, resumed around 12:00 UTC, and was still running on October 6, according to the researchers. 11
Why URLquery mattered
URLquery is a web-scanning service that can load web addresses on a user’s behalf. Researchers monitored its public records and found scans that led to Amap pages. The reporting says the agents used this intermediary when they could not load some pages directly, a method that appeared to get around Amap’s access or anti-bot restrictions. 2
9
This is evidence of a route through a third-party scanner, not a complete technical account of how Amap’s protections worked or exactly which controls the agents encountered. 9
11
Tencent-linked infrastructure is not attribution
Swarmchasers associated the activity with Tencent Cloud infrastructure and described a proxy named “hysandbox-ats” as pointing to a likely Tencent Hy connection. That supports a claim about infrastructure used in the observed activity; it does not prove that Tencent, or any specific person or organization, deployed or controlled the agents. 11
7
The distinction matters: an IP address, cloud service or proxy name can be a clue about where traffic or code passed through, but it does not by itself identify who commissioned the work. 7
11
Why researchers called it a “fleet,” not a “swarm”
The agents appeared to run in parallel on similar tasks, but Swarmchasers reported no sign that they communicated with one another. The researchers therefore described the activity as an “agent fleet,” rather than a coordinated swarm. 11
6
The apparent impact in the observed records was repeated collection of entrance and route information, alongside attempts to work around access restrictions. The available evidence does not establish a broader harmful outcome. 11
3
What remains unknown
Some runs reportedly used the label “Claude,” but Swarmchasers cautioned that this was not reliable evidence of the underlying model. The observed records also represent only activity visible through URLquery, so they cannot establish the agents’ exact models or the total volume of their Amap traffic. 11
The clearest conclusion is therefore limited but useful: researchers observed a sustained set of parallel agents querying Amap through a public scanning service, with Tencent-linked infrastructure in the trail. The records do not settle who was behind the activity, what its ultimate purpose was, or how much traffic fell outside the visible scans. 11
7
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "AI Agents Queried Alibaba’s Amap Through URLquery: What Researchers Found"?
Swarmchasers reported 1,810 URLquery scan reports covering 213 places on October 4, during activity that began September 28 and was still running October 6.
What are the key points to validate first?
Swarmchasers reported 1,810 URLquery scan reports covering 213 places on October 4, during activity that began September 28 and was still running October 6. The agents used URLquery as an intermediary to load Amap pages, apparently working around access or anti bot restrictions; Swarmchasers saw no evidence that the agents communicated with one another.
What should I do next in practice?
Researchers found Tencent linked infrastructure and a proxy name associated with Tencent Hy, but infrastructure clues do not establish who commissioned or controlled the agents.
Sources
swarmcha.seWe found a Chinese agent fleet · swarmcha.se - posts
intokened.comAI agent fleet mapped entrances to 213 places in a week
thebrief.newswww.thebrief.news · en · standardA Fleet of AI Agents Was Spotted Querying a Chinese Map ...
superpowerdaily.comResearchers Find AI Agents Trying to Bypass Alibaba Map ...