Hackers used Claude to automate around 30 attacks.
Anthropic said it is confident the hackers were backed by the Chinese government.
This incident follows a similar one from earlier in the year. At that time, Anthropic reported that Claude had been misused in a “vibe hacking” extortion scheme.
Hackers have once again used Anthropic’s AI model Claude to carry out cyberattacks. The company confirmed to the Wall Street Journal that the Chinese state-backed hackers used the Claude AI model to automate around 30 attacks against corporations and government organisations during a campaign in September. This level of automation, estimated at 80 percent to 90 percent, was significantly higher than in earlier incidents. According to Anthropic’s head of threat intelligence, Jacob Klein, the attacks happened “literally with the click of a button, and then with minimal human interaction.” The human involvement mainly provided quick confirmations or corrections, such as approving actions or asking the model to double-check something.
SurveyThe use of artificial intelligence in hacking is becoming more common across the cybersecurity world. Google recently reported that Russian threat groups have used large-language models to help generate commands for their malware.
Also read: OpenAI officially starts testing group chats in ChatGPT: Here’s what we know
US officials have long warned that China is using AI to steal sensitive information from American companies and citizens, claims China has repeatedly denied. In this most recent case, Anthropic said it is confident the hackers were backed by the Chinese government. The attackers successfully stole sensitive data from four victims, though the company did not name them. Anthropic did note that the US government was not among the successful targets.
Also read: Spotify launches 3 new Premium plans in India: Check pricing and features
This incident follows a similar one from earlier in the year. At that time, Anthropic reported that Claude had been misused in a “vibe hacking” extortion scheme aimed at at least 17 organisations, including groups in healthcare, public services, and government. In that campaign, cybercriminals attempted to demand ransoms of more than $500,000 in exchange for keeping stolen data private.
Also read: Xiaomi 14 Civi price drops by over Rs 18,000 on Amazon: How to grab this deal
Follow Us
Ayushi Jain
Ayushi works as Chief Copy Editor at Digit, covering everything from breaking tech news to in-depth smartphone reviews. Prior to Digit, she was part of the editorial team at IANS. View Full Profile
