Threat Intelligence

One Hacker, a Set of AI Agents and Stolen Taipei Emergency Dispatch Records

The Hacker News · 9 Nov 2026
Key Takeaway Check that any system holding customer or patient data, especially web portals and APIs, requires a login, because AI tools let even a lone attacker find exposed systems at scale.

Researchers at Acronis's Threat Research Unit report that one attacker used AI agents to collect emergency dispatch records covering seven Taipei districts. The agents also found a Taipei fire department platform exposing patient data from 2018, including ECG readings and ambulance video footage, with no login required. The researchers are highly confident the attacker is one person based in Hualien City, Taiwan.

He pointed the agents at hundreds of organisations in at least 11 countries, including emergency services, banks, universities and military systems. Data was collected or systems accessed at about 10 of them in Taiwan, Mexico and Thailand. The evidence came from the attacker's own server, which first looked like a misconfigured development system but was in fact his attack platform. It held more than 90,000 files, including the agents' conversations, his command history, hacking tools, credentials and victims' data.

The agents did not communicate with each other, so he manually pasted one agent's output into another. Commercial AI models mostly refused to help, so the agents ran almost entirely on "abliterated" models: AI run locally with its safety training removed. The dispatch records came from the system behind Taiwan's 119 emergency number and include incident times, response types, case status and street-level locations. They are case records, not live ambulance tracking. The report does not say whether the affected organisations or authorities were told, or whether the exposed data can still be reached. This is based on the opening of the article only.

AI agents data exposure emergency services Taiwan threat research
Running AI agents? Eight controls for securing them, mapped to the ISM ->

Summarised by CISO AI from The Hacker News, written with Claude Sonnet 5.5. We link back to every original so you can read it yourself.