Story perspectives

AI-Powered Attacks: Chinese Hackers Target Corporations, Government

11/14/2025

49 6

Summary

Story summary

Anthropic, the AI safety company, said Chinese state-backed hackers exploited its Claude model to automate about 30 cyberattacks on corporations and government entities in September.
80% to 90% of the attacks were automated, with human operators only confirming critical decisions.
Anthropic confirmed four victims had sensitive data stolen.
The U.S. government was not targeted.
The incident marks a significant escalation in cyber warfare tactics as AI-driven attacks become more prevalent.

1 of 1

Story summary

Anthropic, the AI safety company, said Chinese state-backed hackers exploited its Claude model to automate about 30 cyberattacks on corporations and government entities in September.
80% to 90% of the attacks were automated, with human operators only confirming critical decisions.
Anthropic confirmed four victims had sensitive data stolen.
The U.S. government was not targeted.
The incident marks a significant escalation in cyber warfare tactics as AI-driven attacks become more prevalent.