AI Agents Used in Autonomous Cyberattacks on Taiwan Government Agencies

Taiwan's Ministry of Digital Affairs (MODA) announced on Thursday that overseas hackers used artificial intelligence agents to target government agencies in a series of cyberattacks during July. The intrusion, first detected by the Israeli AI company Dream and reported by the Financial Times, is believed by some experts to be the first known fully autonomous attack on government agencies. The AI system allegedly coordinated up to eight agents to automate reconnaissance, credential attacks, and the strategizing of attack paths without human intervention.

According to reports from Dream, the attackers compromised at least 85 government user accounts and extracted more than 2,500 personnel records. The operation expanded to include Taiwan's nuclear safety agency and at least seven energy companies. MODA stated that the attacks utilized a hybrid approach combining conventional hacking with AI agents such as OpenClaw, which allowed the attackers to combine techniques and exploit secondary systems to make the operation faster and more scalable.

While neither the Taiwanese government nor Dream officially attributed the attack to a specific entity, researchers noted that internal communications linked to the operation were in simplified Chinese. This evidence suggests the operators were likely connected to China. CNN noted that Taiwan frequently faces cyberattacks from China, which claims the island as its own.

Center leaning coverage focused on the official response from the Ministry of Digital Affairs and the technical capabilities of the AI agents. Center left leaning coverage framed the event as a broader warning about the future of cyberwarfare, emphasizing the collapse of the cost for attackers and the urgent need for new defensive strategies and legal frameworks.