Taiwan reports anomalous AI-powered cyberattack
Tengrinews.kz — Taiwan has detected cyberattacks targeting government agencies in which hackers used artificial intelligence-based tools, the Ministry of Digital Affairs said, citing Reuters.
According to the ministry, the attacks originated from abroad. The affected government agencies were able to mitigate the consequences of the incident.
The ministry said cybersecurity specialists detected an “anomalous attack” on government systems in July. Beginning on July 20, Taiwan’s National Institute of Cyber Security started issuing alerts while simultaneously conducting an investigation.
The investigation found signs that the attacks originated outside Taiwan. The hackers used a combined approach: some operations were carried out by humans, while others were performed by AI agents, including OpenClaw.
Note: OpenClaw is an open-source personal AI assistant that runs on a user’s own hardware or server.
“The relevant attack sources, methods, and scope of impact have all been fully investigated, and the affected units have successively completed their handling,” Taiwan’s Ministry of Digital Affairs said.
Following the incident, Taiwanese authorities developed additional recommendations for protection against AI-assisted cyberattacks and stepped up monitoring of government information systems. The Ministry of Digital Affairs did not identify the country from which the attacks originated.
What is known about cyberattacks on Taiwan
In recent years, Taiwan has repeatedly accused China of conducting what it calls hybrid warfare. According to Taipei, this includes military exercises near the island, information campaigns and cyberattacks.
According to Taiwan’s National Security Bureau, the number of cyberattacks targeting the island’s critical infrastructure increased by six percent in 2025 compared with the previous year. An average of around 2.63 million attacks were recorded per day.
In January, the agency also reported that some cyberattacks coincided with Chinese military exercises. Taipei regarded them as part of hybrid threats potentially capable of disrupting critical systems.
The Taiwanese authorities’ statement came one day after Israeli cybersecurity company Dream published the results of an investigation into another AI-assisted hacking campaign.
According to the company, AI agents worked together during the attack to obtain dozens of government officials’ passwords, steal personal data belonging to employees of Taiwan’s Ministry of Justice, and probe the country’s Nuclear Safety Commission for vulnerabilities.
According to Dream, the entire operation took four days. The company said it was able to reconstruct the attack after discovering the complete workspace used by the AI agents.
AI helps hackers accelerate attacks
AI-assisted cyberthreats have existed for several years, but advances in new AI models are allowing attackers to conduct reconnaissance, identify vulnerabilities and prepare attacks significantly faster.
Semgrep researcher Cris Thomas said the incident demonstrates how quickly malicious actors can use AI to achieve their objectives.
At the same time, he cautioned against overstating the autonomy of such systems.
“There's still a human in there somewhere. Somebody had to choose who to attack, had to establish an objective and give it a directive. It's not totally 100% autonomous. There was a capable operator in charge that did that,” Thomas said.
According to him, a skilled operator was behind the attack, determining both the target and the direction of the operation.