AI-Assisted Cyberattack Targets Taiwan’s Government Agencies

Taiwan’s Ministry of Digital Affairs announced that its cybersecurity monitoring units identified an “abnormal attack” against multiple government agencies in July. Beginning on July 20, the National Institute of Cyber Security issued warning alerts while investigating. The ministry said the attacks combined manual operations with AI-agent assistance—specifically tools such as “Open Claw”—and that the affected bodies have “successively completed their handling.”

Reuters

Taiwan has repeatedly warned that China employs “hybrid warfare,” mixing military drills, disinformation and cyber operations to pressure the island. The National Security Bureau reported that cyberattacks on Taiwan’s critical infrastructure—including hospitals and banks—rose 6 % in 2025 compared with the prior year, reaching an average of 2.63 million attacks per day. Some of those incidents were timed to coincide with Chinese military exercises, a pattern the bureau described as “hybrid threats” intended to paralyze the island.

Reuters

The Ministry of Digital Affairs said it has introduced protective guidelines and strengthened system monitoring across agencies to block future AI-derived threats. The statement did not name any foreign actor. An Israeli cybersecurity firm, Dream, posted a blog describing an AI-driven hacking campaign that stole credentials and scanned a nuclear safety agency over four days. Dream reconstructed the attackers’ “complete operational workspace” but declined to disclose the data or name the targeted government. The Financial Times, which first received Dream’s briefing, identified the compromised entities as Taiwanese agencies, including the justice ministry and the nuclear safety agency.

Reuters

According to Dream’s analysis, a coordinated team of AI agents extracted large numbers of passwords from unidentified officials, pilfered personnel records from Taiwan’s justice ministry, and probed the nuclear safety agency for vulnerabilities. The attacks were executed over a four-day period and displayed clear characteristics of an overseas source, according to the Ministry’s investigation.

Reuters

The incident underscores the growing sophistication of state-linked cyber operations that blend AI automation with human direction. Taiwan’s rapid response—issuing alerts, completing remediation, and updating defensive protocols—illustrates an evolving defensive posture in a region where cyber and kinetic threats are increasingly intertwined. The episode also highlights the challenge for governments to attribute AI-enhanced attacks, as official statements stopped short of naming a perpetrator while industry analysts linked the campaign to broader “hybrid warfare” dynamics.

Reuters