Autonomous hacking
I've tested an autonomous hacker on real products over the past month,[1] and it's very capable. Concretely, I could: 1. Hijack accounts at a major bank and log in as other users. 2. Bypass authorization in a major AI lab's product and access other users' private data, including uploaded
I've tested an autonomous hacker on real products over the past month, [1] and it's very capable. Concretely, I could: Hijack accounts at a major bank and log in as other users. Bypass authorization in a major AI lab's product and access other users' private data, including uploaded files. List all users of a big tech company's product and modify their files. Download the health records of anyone whose data was stored and sharable through a popular electronic health records system. [2] If I were malicious, I reasonably believe that I could cause billions of dollars in damage in less than a mon
related reading
- The End-State Fallacy: Where Is AI Security Headed?endstatefallacy.com
- Dario Amodei — The Adolescence of Technologydarioamodei.com
- Countering misuse of AI: September 2026 / Anthropicanthropic.com
- Project Glasswing: Securing critical software for the AI era \ Anthropicanthropic.com
- Training a Misaligned Reward Seekeralignment.anthropic.com
- Concrete Evaluations to Investigate the OpenAI Model That Hacked Hugging Face — LessWronglesswrong.com
- Greg Brockman on Svbtleblog.gregbrockman.com
- Vulnerability Research Is Cooked - Quarrelsomesockpuppet.org
- The OpenAI/Hugging Face Incident: Challenges in Controlling and Containing Cyber-Capable AI Systems - Institute for AI Policy and Strategyiaps.ai
- Announcing Safety Research Grantsthinkingmachines.ai
- A Safe Path to Open Weights - Thinking Machines Labthinkingmachines.ai
- OpenAI – Hugging Face Incident Technical Reportcdn.openai.com