Skip to content
iThome (Taiwan)· 張明德·Indexed by this site · Originally published· Event →· Its place in the scan record →

Shadowserver observation: 3,847 instances of AI attack frameworks exposed on the public internet, 20 in Taiwan

公開網路上存在逾3,800個AI攻擊框架執行個體,臺灣也有20個

AI summary

Security monitoring organization Shadowserver Foundation found in its tracking that exposed instances of AI autonomous attack frameworks on the public internet involved a total of 3,847 distinct IPs as of November 15, including 20 in Taiwan. Previously, the FBI issued a warning on November 13, saying that hackers had been using the open-source AI autonomous penetration testing framework ARTEX to launch attacks at least since August of this year, greatly accelerating intrusion speed, after which Shadowserver began daily tracking of exposed instances of frameworks such as ARTEX, CyberStrike, PentestCode, and VXControl. Among the frameworks, ARTEX had 2,852 IPs, CyberStrike had 493, PentestCode had 231, and VXControl had 196; geographically, the United States had the most with about 1,400, followed by China with 637, Hong Kong with 624, Singapore with 163, and Japan with 147. However, the location of an IP is not equivalent to the location of the operator, nor does it mean that these hosts are all being used for attacks.

Read originalithome.com.tw