Summary
The AI company Anthropic claims that Chinese government hackers used its chatbot, Claude, to automate cyber attacks on about 30 global organizations. The hackers allegedly tricked the chatbot by pretending to be cybersecurity researchers, using it to perform small tasks that added up to a sophisticated cyber espionage campaign. Anthropic has informed affected companies and authorities while identifying that the hackers were possibly state-sponsored.
Key Facts
- Anthropic is an AI company that created the chatbot Claude.
- Anthropic claims hackers used Claude in a cyber-espionage campaign.
- Hackers posed as cybersecurity researchers to trick the chatbot.
- Around 30 global entities were targeted, including tech companies and government agencies.
- Hackers built a program using Claude's help to compromise targets autonomously.
- Anthropic has banned the hackers and notified affected organizations and authorities.
- There is debate over the effectiveness and claims of AI-driven cyber attacks.
- Other AI firms, like OpenAI, have also reported state-affiliated actors misusing their tools.
