dfcheng··on We Found Zero Low-Severity Bugs in 165 AI Code Reports. ZeroThis is what I’ve experienced having LLMs code: ensuring security is not an adequate part of its training. Of course, modern developers I work with don’t give a shit either.