Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Clear out junk files and repair common Windows errors3Scan for outdated or missing drivers - takes under a minuteResearchers have documented parallel programs querying Alibaba’s Amap mapping service for entrance-navigation information about public places in China. Their preliminary report links the observed request path to Tencent Cloud infrastructure, but it does not identify who directed the activity, why it was carried out, or which AI model was used.
What did the researchers observe?
In a preliminary report dated October 4, 2026, and updated October 5–6, the independent group Swarmchasers described programs submitting code to urlquery.net, a domain-scanning service. Its browser then loaded Amap pages and APIs. The researchers interpreted the requests as attempts to gather entrance-level navigation shares for public locations, including parks, museums, zoos, and hospitals. TechCrunch’s October 5 account also described the activity and said the investigation was ongoing.
As an Amazon Associate I earn from qualifying purchases.
The records show requests and resulting reports, not a confirmed account of what an operator intended to do with the information. The report is preliminary; the researchers said a fuller report would follow.
Why call it an “agent fleet” instead of a “swarm”?
The researchers saw parallel runs performing similar queries, but found no sign in the records they reviewed that the runs communicated with one another. They therefore used “agent fleet” rather than “swarm.” The phrase describes their interpretation of the observed behavior; it does not prove a particular system design or that every run acted independently.
#1 Best Overall
Swarmchasers counted 2,048 Amap reports across 216 places from September 28 through October 4, 2026, including 1,810 reports on October 4 alone. They reported up to 14 runs active at once and 51 places worked on in the busiest hour. Exact simultaneity was rare, the researchers cautioned, and a handful of fast agents could account for the pattern rather than a large number of independent agents.
Those totals are based on the researchers’ collection and classification of urlquery records. The report says records expire, making the counts lower bounds; labels do not reliably identify runs, and urlquery does not store page content.
What did the entrance-share results say?
By the report’s update, the researchers had five entrance-share readouts. One documented readout for Chengdu Zoo assigned 71% to the North Gate, 23% to the East Gate, and 6% to the Southeast Gate. Another location’s readout listed Ground Car Park at 40%, Main Gate at 26%, and smaller shares for other entrances.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteThese are figures reported by the observed Amap activity, not independently verified measurements of actual visitor behavior. The report does not establish how the percentages were produced or whether they reflect real-world traffic.
Rank #3
What does the Tencent Cloud connection establish?
Swarmchasers connected observed requests to Tencent Cloud using inbox creation records and requests from Tencent Cloud addresses with a proxy Via header ending in “hysandbox-ats.” This supports a link between the request path and Tencent Cloud infrastructure. It does not show who controlled the infrastructure or directed the agents: Tencent Cloud is available to outside customers, and the researchers note that a Via name is self-reported.
The researchers also said their tests did not match Tencent’s publicly described Agent Sandbox product in its standard internet mode. They had not reviewed the data behind those tests, so that comparison does not identify the system used.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Was Claude or another model behind the activity?
The model has not been established. Although 211 reports carried a “claude” label, the researchers said code-style comparisons and limited model tests appeared more consistent with Tencent Hy and other Chinese model families than with Claude. No record they found named a model or training job.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →The report cautions that its tests were small and that model self-identification can vary with the prompt and language. Neither the labels nor the comparisons prove which model, if any, powered the observed runs.
Best Value
What remains unknown?
- Operator: The report does not identify the person or organization directing the activity.
- Purpose: The records do not distinguish an evaluation or task-generation run from training rollouts, and do not establish how the entrance information was meant to be used.
- Model: No definitive model or training job is named.
The available account documents activity through October 6, 2026. It supports describing Amap queries and a Tencent Cloud infrastructure connection, not calling the activity espionage, an attack, or a confirmed government operation.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
