Daily News

Anthropic Accuses China-Supported Hackers of Targeting 30 Organizations

Published

on

U.S. artificial intelligence company Anthropic has stated that a hacker group backed by the Chinese government used its AI technology to launch a highly automated cyberattack on dozens of tech companies and government agencies. The incident occurred in September and is considered the first publicly disclosed case of AI agents conducting large-scale intelligence collection with minimal human intervention.

In its latest report, Anthropic stated that the attackers utilized its generative AI tools to write code and directed its AI agent, “Claude Code,” to execute the intrusion operations. Only about 10–20% of the tasks required direct human control, while the AI autonomously executed the remainder. Although Anthropic did not disclose the names of the 30 targeted entities or the methods used to identify the hackers, the company said it was “highly confident” that the operation was led by a cyberattack group supported by the Chinese government.

In response to the allegation, Chinese Foreign Ministry spokesperson Lin Jian said that China had not seen the report and criticized the U.S. company for making “baseless accusations without evidence,” reiterating that China consistently opposes any form of hacking.

This is not the first warning from AI developers about the misuse of their technology. Earlier this year, Microsoft and OpenAI also reported that government-backed actors were using AI to enhance cyber intrusions and surveillance operations. In its latest annual threat report, Microsoft identified China, Russia, Iran, and North Korea as expanding the use of AI for influence operations and cyberattacks. In February, OpenAI also reported that Chinese security agencies were using AI tools to monitor anti-China content on Western social media platforms.

Since August, Anthropic has disclosed that Chinese hackers used its technology to target Vietnamese government databases and telecom companies. In September, the company updated its service terms to further restrict access to its AI tools in certain regions, explicitly naming China as the only country affected.

Trending

Copyright © 2021 Blessing CALD