Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Why AI-generated vulnerability patches still require expert human review (1password.com)
3 points by wrxd 28 days ago | hide | past | favorite | 1 comment


LLMs produce Fix-Like Artifacts with Embedded Defects (FLAWED) 53.9% of the time

A human that produced similar results would likely be looking for a different job soon.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: