A Security Company Says Government-Linked Hackers From China Used AI Programs to Attack More Than 200 Organizations in 11 Days
The security company CrowdStrike says a spying group linked to China used a team of AI programs, steered by a person, to go after more than 200 governments and businesses in at least 14 countries in under two weeks, getting inside at least nine. Until now, cases like this involved AI helping with one break-in at a time. CrowdStrike says this is the first time it has seen one operator use AI to run many at once.
CrowdStrike, one of the largest computer security companies, published a report today about a spying group it calls VAPOR PANDA. CrowdStrike has tracked the group since 2021, links it to China, and says it usually goes after governments, research groups and energy companies.
According to the report, starting on October 17 the group used a free, openly published hacking tool called ARTEX. ARTEX runs a team of AI programs: one in charge, others that plan, and others that do the work, all sharing notes about what they have found. The programs ran mainly on DeepSeek, an AI program made in China, plus a second paid AI service CrowdStrike could not confirm. A person stayed in charge, writing instructions in Chinese and stepping in at key moments, but the AI programs did much of the searching for weak spots on their own, trying several ways in at once and switching approach when one failed.
CrowdStrike says the tool was used against more than 200 organizations in at least 14 countries over 11 days. Targets included foreign ministries in Latin America, Europe, Asia and the Middle East; a European country's president's and prime minister's offices; and American companies working on medicine, computer chips, defense and nuclear power. CrowdStrike confirmed the attackers got inside at least nine organizations, including phone companies in the United States, Belgium and the Philippines, and planted a hidden way back in at some of them. At least 12 more faced attempts to break in by guessing passwords.
What's new, CrowdStrike says, is scale. Earlier cases showed AI helping attackers with one break-in at a time. Here one operator ran many attacks at the same moment. CrowdStrike expects groups linked to China to keep adopting tools like this over the next year. It also says the group's goals haven't changed; the AI let it go after more targets faster.
Some caution is warranted. This is one company's account, built partly from material the attackers left exposed on the internet, and no outside researchers have checked it yet. CrowdStrike marks several findings as "likely" rather than certain, and the "more than 200" counts organizations targeted, not organizations broken into. Earlier this month, a separate report described criminals using a different free tool to run AI programs against company websites; this one points to government spies doing the same.
Sources
- CrowdStrike — VAPOR PANDA uses AI-driven ARTEX to scale a global espionage campaign (threat report)crowdstrike.com · primary
- NBC News — China-linked hackers used AI to target 200 organizations, CrowdStrike saysnbcnews.com
- Security Boulevard — Chinese hackers use DeepSeek AI to target more than 200 organizationssecurityboulevard.com