Taiwan says government agencies targeted in AI-assisted cyberattack from overseas

Aug 13, 2026

Taiwan's Ministry of Digital Affairs revealed that government agencies were targeted in July 2026 by an overseas cyberattack utilizing open-source AI agents. Cybersecurity firm Dream reported the campaign compromised 85 accounts, stole 2,500 justice ministry records, and scanned nuclear and energy infrastructure. While Taiwan did not directly name China, linguistic evidence strongly links the operators to Beijing, which continues to exert hybrid warfare pressure on Taipei.

AI-assisted cyberattacks on Taiwan

▪Taipei's National Institute of Cyber Security issued warning alerts beginning July 20, 2026, as authorities investigated the abnormal cyberattack targeting government agencies.
▪The July 2026 AI-driven cyberattack scanned Taiwan's nuclear safety agency for vulnerabilities and targeted at least seven energy companies.
▪Israeli cybersecurity firm Dream reported that an AI-driven hacking campaign compromised at least 85 government accounts and stole over 2,500 personnel records from Taiwan's justice ministry.
▪Taiwan's Ministry of Digital Affairs detected an overseas, AI-assisted cyberattack targeting government agencies in July 2026, which affected units successfully handled.

Chinese hybrid warfare tactics

▪Taiwan's National Security Bureau described coordinated Chinese cyberattacks as part of hybrid threats aimed at potentially disrupting or paralyzing the island, sometimes occurring alongside military drills.
▪Taiwan's National Security Bureau reported that Chinese cyberattacks targeting key infrastructure, including hospitals and banks, rose 6% in 2025 to an average of 2.63 million attacks a day.

Open-source AI agent methodology

▪The July 2026 cyberattack on Taiwan used a hybrid method combining manual operations with open-source AI agent frameworks, including OpenClaw and Hermes.
▪The AI hacking tool bypassed security measures by disguising the intrusion as a sanctioned security test and autonomously re-routed its attack path when blocked by defenders.
▪Israeli cybersecurity firm Dream reconstructed the July 2026 hacking campaign after recovering a 160MB archive of 1,395 files containing the AI agents' complete operational workspace left exposed online.

Taiwan cybersecurity protective measures

▪Amir Becker of Dream stated that the incident represents the first known end-to-end autonomous AI attack on a government, warning that governments must treat themselves as permanently under assault.
▪In response to the July 2026 AI-assisted cyberattack, Taiwan's government introduced new protective guidelines and strengthened cybersecurity monitoring across its agencies.

Cross-strait sovereignty dispute

▪Internal communications tied to the July 2026 cyberattack used Simplified Chinese, leading cybersecurity firm Dream to state it was highly likely the operator was linked to China.
▪Beijing claims sovereignty over Taiwan and has stepped up military and political pressure, including hybrid warfare tactics, to force Taipei to accept its claims.