Taiwan detected AI-assisted cyberattacks targeting government agencies last month, with officials saying the affected bodies successfully handled the incidents.
The Ministry of Digital Affairs said its cybersecurity monitoring units detected an “abnormal attack” against government agencies in July.
Taiwan’s National Institute of Cyber Security began issuing warning alerts on July 20 while investigating the activity.
Overseas Source Identified
The ministry said its investigation found clear signs that the attacks originated overseas.
According to the ministry, the hackers used a hybrid method that combined manual operations with AI-assisted agents, including a tool identified as OpenClaw.
PSG Paris Saint-Germain Retain UEFA Super Cup with 2-1 Win Over Aston Villa
Officials said they investigated the attack sources, methods and scope of the incidents.
The affected government agencies subsequently completed their response measures, the ministry said.
Taiwan has now introduced additional security guidelines and strengthened monitoring across government systems to detect and block similar attacks at an early stage.
Taiwan Faces Growing Cyber Threats
Taiwan has repeatedly accused China of using what it describes as hybrid warfare, including military pressure, disinformation campaigns and cyberattacks.
The island’s National Security Bureau said in January that cyberattacks against key Taiwanese infrastructure increased by 6% in 2025.
The attacks reportedly averaged 2.63 million attempts a day, targeting sectors including hospitals and banks.
The bureau also said some cyber activity coincided with Chinese military exercises, describing the pattern as part of broader hybrid threats.
Israeli Cybersecurity Firm Reports Similar Campaign
Taiwan’s announcement came a day after Israeli cybersecurity company Dream reported an AI-driven hacking campaign targeting an unnamed Asian government.
Dream said a group of AI agents worked together to obtain passwords, steal personnel records and scan a nuclear safety agency for vulnerabilities over four days.
The company did not identify the government or release the underlying data when approached by Reuters.
However, a source identified the agencies involved as Taiwanese.
AI-Assisted Hacking Raises Concerns
Security researchers say AI-assisted cyber operations are becoming more sophisticated.
Cris Thomas, a security researcher and security advocate at code security company Semgrep, said the reported campaign demonstrated how quickly AI could help attackers conduct operations.
However, he cautioned against describing such campaigns as fully autonomous.
Thomas said human operators still decide which targets to attack, establish objectives and direct AI systems.
