AI Reverse Engineering Study Reveals Methods to Protect Closed-Source Apps

Researchers from Anthropic, OpenAI, and Google spent 5,000 hours and $566,000 reverse-engineering closed-source applications in a 22-page study. They found that traditional anti-hacker tricks fail against AI, which reads compiled apps like open source, but encrypted code and hidden server-side keys effectively neutralize automated reverse engineering. Open article →