AI Agents from Tencent Cloud Target Alibaba’s Amap in Data Harvesting Attempts

The report emphasized the growing challenge of distinguishing legitimate AI operations from harmful data harvesting

Researchers identified over 2,600 data extraction attempts by AI agents operating on Tencent Cloud against Amap, Alibaba’s mapping platform, between September 28 and October 8. The activity was uncovered by independent analysts monitoring automated agent behavior across online services. Initial findings were shared on October 4, with a full report released ten days later detailing the scale and origin of the probes. The agents executed repeated requests to gather geographic and location-based data, suggesting a coordinated effort to scrape Amap’s public interface. Researchers noted the pattern resembled automated reconnaissance, though no breach of Amap’s core systems was confirmed. Tencent Cloud’s infrastructure was traced as the host environment, raising questions about the use of its resources for such activities.

The report emphasized the growing challenge of distinguishing legitimate AI operations from harmful data harvesting. How the Activity Was Detected and Attributed Researchers used network traffic analysis and behavioral fingerprinting to isolate the AI agents, focusing on request frequency, timing, and headers inconsistent with normal user behavior. By correlating IP ranges with known Tencent Cloud allocations, they linked the activity to specific virtual instances. The team stated the agents mimicked legitimate API calls but at volumes far exceeding typical usage, triggering alerts in their monitoring systems. They clarified that attribution to Tencent Cloud does not imply direct involvement by Tencent, as cloud resources can be leased by third parties. What Motivates Such Large-Scale Data Collection Efforts? The scale of the attempts suggests potential use in training machine learning models, enhancing competitor mapping services, or building proprietary location databases.

Researchers speculated the data could support applications ranging from autonomous navigation to urban planning tools

Researchers speculated the data could support applications ranging from autonomous navigation to urban planning tools, though the end purpose remains unconfirmed. They warned that unrestricted scraping risks overwhelming service infrastructure and violating terms of use, even if data is publicly accessible. The incident highlights tensions between open data access and protection against automated exploitation. Frequently Asked Questions Did Amap suffer a security breach from these AI agent activities? No, the researchers found no evidence of unauthorized access to Amap’s internal systems or data stores. The attempts were limited to probing public-facing interfaces, resembling aggressive data scraping rather than a cyber intrusion. Could the AI agents have been operated by Tencent itself? The report does not accuse Tencent of direct involvement. It notes the agents ran on Tencent Cloud infrastructure, which could be used by external customers, making attribution to the company uncertain without further evidence.

What actions have been taken since the report’s release? Neither Alibaba nor Tencent has publicly responded to the findings as of the report’s publication date. Researchers recommend improved rate limiting and bot detection mechanisms for mapping services to mitigate similar activity.