I'm assuming they still use AI to find vulnerabilities, just not fix them?