Taiwan Confirms AI-Powered Cyberattack on Government Agencies

Taiwan's Digital Development Ministry has confirmed a cyberattack targeting government agencies, stating that the intrusion was detected in July and has since been investigated and addressed. The attack, reportedly carried out by suspected Chinese hackers, leveraged publicly available Artificial Intelligence tools to autonomously probe government systems.

According to the ministry, the attackers used open-source AI agents to create a tool that could operate like a coordinated hacking team. This autonomous tool was deployed in early July, with multiple agents probing 21 government systems within four days. The AI agents could independently identify vulnerabilities and adjust their strategies when encountering system defenses, highlighting a new frontier in cyber warfare.

The ministry confirmed that the attack exhibited characteristics of being of foreign origin, with hackers employing a hybrid approach combining traditional methods with AI agent assistance. These AI agents can rapidly link various attack techniques and use secondary systems, such as backup or testing environments, as stepping stones. This allows for attacks that are fast, low-cost, and large-scale.

In response to these evolving threats, Taiwan's government has established protective guidelines and enhanced system monitoring across agencies. The Digital Development Ministry emphasized that the investigation into this specific incident is complete, and affected units have finished their processing and repairs. The ministry is committed to continuously improving the nation's overall defense capabilities to counter new forms of cyber threats driven by rapid AI development.