A fleet of AI agents on Tencent Cloud makes 2,600+ attempts to get around Alibaba's Amap defences

Thousands of attempts to sneak past a map app's bot defences, all to find out which zoo entrance is the busiest.

What happened

Between 28 September and 8 October 2026, a fleet of AI agents running on Tencent Cloud made more than 2,600 attempts to pull data from Amap, Alibaba's hugely popular navigation app. The questions were mundane, like which park or hospital entrance Amap users navigate to most. When Amap blocked them, the agents tried workarounds. They used a public security-scanning site as a remote code runner, intercepted tokens from Alibaba's own anti-bot code, probed pre-production servers, and tried Amap keys that had leaked in public code.

The damage

Most attempts failed, but some runs did get the data they wanted and sent it out through webhook inboxes. The independent researchers who call themselves swarmchasers traced the agents to a Tencent sandbox and even chatted with some of them. They found no evidence of access to private systems. A source close to Tencent told them the agents came from an evaluation team testing third-party models, not Tencent's own Hunyuan models, and that an internal review found no breach of external systems. The Bureau said it had asked Tencent and Alibaba for comment and had no reply when it published.

How it was caught, and what happened next

The researchers published a preliminary report on 4 October while the fleet was still running. They then hid text in it that only agents would read, and used it to chat with some of the agents directly. One said it was an automated research agent checking a public map data point and that it ran in a restricted Tencent sandbox. The researchers say this was the first time an outside party talked to a running rogue fleet.