Chinese hackers used Claude artificial intelligence to attack

Anthropic announced Thursday that government-sponsored hackers chinain September of the artificial intelligence model Claude This company to automate approx 30 attacks They have been used by companies and governments.

Hackers misuse of Claude artificial intelligence

According to the Wall Street Journal report, Entropic said that 80 until 90 percent from attacking automatically with artificial intelligence It has been done, and its amount has increased compared to previous similar attacks. About the attacks, Jacob Klein, head of threat intelligence at Anthropic, explained that they are “virtually done with one click and then with minimal human interaction.”

He said in this regard:

“The user intervened at only a few critical points and was asked commands such as “Yes, continue,” “Don’t continue,” “Thanks for the information,” “Oh, that doesn’t sound right,” and “Claude, are you sure?” has been using.”

Anthropic told the Wall Street Journal that it believes the hackers were sponsored by the Chinese government. In this campaign, the hackers stole sensitive data from four victims, but as with previous attacks, Anthropic did not disclose the names of the targets and whether the attack was successful or not.

Hacking using artificial intelligence models has become very common in today’s world and new strategies are being developed for it. Recently, Google identified Russian hackers who were using large language models to generate the commands required for their malware.

For several years, the US government has been warning that China is using artificial intelligence to steal data from American citizens and companies. A claim that China has denied.

Anthropic recently published a report in which a technique called “vibe hacking” was mentioned. The company said in its report that agent-based artificial intelligence systems are now becoming tools for cyber attacks and can carry out complex operations.