Linus Torvalds recently used AI to help debug a Linux kernel bug, crediting it with "much of the grunt-work" while noting the generated code still needed oversight. The article argues LLMs are more effective at finding bugs than fixing them, pointing to maintainers being flooded with AI-generated bug reports and fixes, some escalated through emergency security channels. It also suggests bug-finding use may sidestep plagiarism and copyright concerns behind some distros' AI bans.
No score is assigned. Sources and their independence are shown in the citation chain below.