Taiwan Confirms AI-Assisted Cyberattacks on Government Agencies

Taiwan reports first confirmed AI-assisted government hack

Taiwan's Ministry of Digital Affairs confirmed Thursday that government agencies successfully defended against AI-assisted cyberattacks detected last month, marking one of the first publicly acknowledged incidents of automated hacking tools targeting a national government.

The attacks, which began July 20, employed what the ministry described as a "hybrid approach" combining manual hacker operations with AI agent-assisted techniques. The ministry specifically identified Open Claw among the tools used in the campaign, which originated from overseas sources.

Cybersecurity monitoring units issued multiple warning alerts while investigating the incident. According to the ministry's statement, all affected agencies have completed their response procedures, and the government has established new protective guidelines to counter AI-derived cybersecurity threats.

Why it matters

This incident represents a significant escalation in the automation of state-level cyberattacks. While AI-assisted hacking tools have existed for years, their deployment against government infrastructure demonstrates how rapidly these capabilities are being operationalized. Taiwan faces an average of 2.63 million cyberattacks daily—a 6% increase from 2024 to 2025—making it a leading indicator for threats other nations will likely face as AI hacking tools proliferate.

Israeli firm reveals operational details

The Taiwanese government's disclosure came one day after Israeli cybersecurity company Dream published findings about an AI-driven hacking campaign targeting an unnamed Asian government. Dream reported that a team of AI agents worked collaboratively over four days to extract passwords from officials, steal personnel records from Taiwan's justice ministry, and scan its nuclear safety agency for vulnerabilities.

Dream reconstructed the attack after recovering what it described as the agents' "complete operational workspace," though the company declined to publicly share the data or confirm the target government's identity. The Financial Times first reported that the targeted agencies were Taiwanese.

Growing automation in cyber warfare

The threat landscape shifted dramatically in recent months following the release of advanced AI models capable of conducting reconnaissance, identifying vulnerabilities, and exploiting them with minimal human oversight. Anthropic's Mythos model represents one such tool that has accelerated the pace of AI-assisted attacks.

Cris Thomas, a security researcher at code security company Semgrep, cautioned against overstating AI autonomy in these operations. "There's still a human in there somewhere," Thomas noted. "Somebody had to choose who to attack, had to establish an objective and give it a directive. It's not totally 100% autonomous."

Taiwan has characterized the increasing frequency of cyberattacks as part of China's "hybrid warfare" strategy, which combines military drills, disinformation campaigns, and digital intrusions. The island's National Security Bureau reported in January that some cyberattacks are synchronized with military exercises to create coordinated pressure.

The Ministry of Digital Affairs did not explicitly attribute the July attacks to any specific nation, and China's Taiwan Affairs Office did not respond to requests for comment.

These details were first reported by Reuters correspondents Ben Blanchard and Raphael Satter.