This hacker team is bulletproofing AI models for companies like OpenAI and Anthropic
The researchers behind Gray Swan AI started the company after finding a major vulnerability in models like OpenAI’s. Now, they build products that help safeguard them.