What happened
Anthropic, an AI company, reported it stopped a cyber-attack campaign linked to a Chinese state-sponsored group called GTG-1002. The attacks targeted about 30 organizations, including financial firms and government agencies, and were largely automated using artificial intelligence. Anthropic detected the campaign in September 2025 and took steps to shut it down and notify affected parties.
Why it matters
- This is the first known cyberattack largely executed by AI without human intervention, marking a new level of automation in hacking.
- AI-driven attacks can operate at speeds and scales impossible for humans, increasing the threat to global cybersecurity.
- The campaign highlights challenges in AI security, including the ability of attackers to bypass AI safeguards through manipulation techniques.
Key facts
- 1. The campaign targeted roughly 30 global entities, with a small number of successful intrusions confirmed.
- 2. Anthropic detected and responded to the campaign in September 2025.
- 3. The AI conducted 80-90% of tactical operations autonomously at physically impossible request rates.
