Anthropic: "Chinese Hackers Used AI Claude to Attack 30 Government and Corporate Targets"
Anthropic revealed on November 13 (local time) that state-sponsored hackers backed by the Chinese government used Anthropic's artificial intelligence (AI) technology last September to carry out cyberattacks against major corporations and foreign governments in an "automated manner."
The company stated that the attack targeted dozens of entities and involved the highest level of automation ever observed in a hacking incident. The hackers attempted to infiltrate approximately 30 companies and government agencies. However, Anthropic did not specify which companies or countries were targeted. The company also confirmed that some sensitive information was exposed during the process.
Anthropic is convinced that the attack was carried out by hackers supported by the Chinese government. However, the Chinese Embassy in Washington did not comment on the incident. The Chinese government has consistently denied involvement in hacking cases in which it has been implicated.
AI Performed 80-90% of the Tasks..."Humans Only Needed to Click 'Yes' or 'Continue'"
Anthropic reported that this hacking incident was the most advanced AI-based automated attack it has witnessed to date. In the past, hackers wrote code themselves, hunted for system vulnerabilities, and crafted phishing emails manually. In this case, however, AI performed 80-90% of the entire operation. Human intervention was limited to a few critical decision-making steps.
Jacob Klein, Head of Threat Intelligence at Anthropic, stated that "the hackers could literally execute the attack with the press of a single button, and there was almost no human involvement afterward." He added, "The hacker's role was limited to reviewing a few things, such as clicking 'yes,' 'continue,' 'thanks for this information,' or asking 'Claude, are you sure about this? This seems odd.'"
The Wall Street Journal (WSJ) pointed out that cyberattacks are evolving in a completely different direction through AI-based automation. Most of the work that previously required human labor and technical skills from hackers can now be performed autonomously by AI.
Recent hacking cases also confirm that cyberattacks are now being automated through the use of AI. This summer, cybersecurity company Volexity detected a case in which Chinese hackers used AI tools to automate target selection, phishing email composition, and even malware creation. In another recent incident, hackers linked to the Russian government used AI models to generate customized malware instructions in real time during attacks on Ukraine.
The growing trend of cyberattacks using AI agents highlights the dual nature of AI tools. The WSJ noted, "Anthropic is committed to using AI to strengthen cyber defense, but more powerful AI can also become a more powerful means of attack."
-
Reporter Kim Minyoung argus@asiae.co.kr
-
Edit Hong Jayeon hongzaa@asiae.co.kr
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
