Kathrine Frich
The files showed the attack used open-source systems known as Hermes and OpenClaw.
Modern tools are designed to make our lives easier, but they can quickly slip into the wrong hands.
Tracing digital footprints gets complicated when an invisible intruder breaks in.
Now, a new kind of threat is reshaping how nations protect their secrets.
An invisible digital team
Suspected hackers from China recently used public artificial intelligence to breach government websites in Taiwan. The Financial Times cited by United24Media reports the operation marked a frightening milestone in digital warfare.
Software agents handled the job completely on their own, shifting tactics whenever they hit a wall.
The operation ran over four days in early July. Up to eight automated agents mapped out 21 government systems and hunted down weak spots.
Stealing the data
The software compromised at least 85 government accounts. It grabbed more than 2,500 personnel files before targeting Taiwan’s nuclear safety agency and several energy companies.
An Israeli cyberdefense firm named Dream first uncovered the breach. While tracking online threats, analysts found the hidden evidence inside a massive online archive.
The files showed the attack used open-source systems known as Hermes and OpenClaw. To bypass standard safeguards, hackers tricked the software by claiming they were running an authorized security test.
When one digital path failed, the system dispatched another agent. The tool constantly reorganized its attack plans using fresh intelligence.
A global warning
Experts say this marks a massive shift. Amir Becker, chief strategy officer at Dream, called the breach an unprecedented “end-to-end autonomous attack” against a government target.
He issued a stark warning for world leaders. “This must be the basic assumption of every government around the globe,” Becker told the Financial Times.
Dream did not specifically name Taiwan or China, but other sources quickly confirmed the target. Analysts also noticed Simplified Chinese in the internal communications, a detail pointing strongly toward Beijing.
The new normal
A spokesperson for Taiwan’s Ministry of Digital Affairs declined to discuss the specific case, but acknowledged the growing danger.
“Hacker attack methods have become increasingly diverse, and in recent years, the integration of AI technology has further transformed the nature of cyber security incidents,” the spokesperson noted.
Taiwan faces a relentless digital barrage. The island’s National Security Bureau reported an average of 2.6 million Chinese cyberattacks every day in 2025.
Last year, Anthropic reported that Chinese groups manipulated its software to breach dozens of global companies.
Sources: Financial Times, United24Media, Taiwan’s Ministry of Digital Affairs, Taiwan’s National Security Bureau, Anthropic
