they stole Claude’s brain 16 million times

The most safety obsessed AI company on the planet just watched their own AI get weaponized into an autonomous hacking machine. Here's what happened. A Chinese state sponsored group called GTG 10002 took Anthropics Claude, the AI built to be the safest in the world, and turned it into an autonomous hacking engine. It hit about 30 organizations, tech companies, banks, government agencies. How'd they do that? They told Claude it was doing authorized defensive security testing. That's it. No fancy exploit. They just lied to it, which is like 90% of AI hacking. And Claude went to work. recon, vulnerability scanning, custom exploit code, credential harvesting, data extraction. 80 to 90% of the campaign ran without a human touching it. Thousands of requests per second. There were only about four to six human decisions in the entire operation. Anthropic themselves have said that the barrier to performing these sophisticated cyber attacks has dropped substantially. One AI can now do the work of entire teams of experienced hackers. So yeah, the world's biggest AI safety company just got hacked by their own AI. And now less experienced groups can launch nation state level attacks. That's where we are. Fun.