Corporate Linux

Recent kernel decisions push Linux toward a corporate, AI-driven model: maintainers will remove mandatory "Assisted-by" tags while asking contributors to disclose creation methods in cover letters instead of changelogs. That shift hides whether patches were generated by LLMs, making it difficult to trace defects to AI output. U.S. Copyright Office guidance requires authors to identify human-authored versus AI-generated portions to claim copyright; concealing which parts are AI-generated undermines enforceable copyright, weakens the GPL’s protections, and clears a legal path for companies to keep LLM-assisted kernels private. Kernel developers acknowledge companies will maintain private, AI-aided kernels and appear willing to accept that outcome.

Linus Torvalds’ embrace of LLMs accelerates the trend: he uses a heavily tuned LLM workflow for multi-hour cleanup passes that he says require paid resources and months of setup, and he refuses to share his configuration. He explicitly favors limiting contributions to "serious" developers with resources and endorses making it harder for newcomers to contribute. That stance follows a long pattern of privileging pragmatic commercial outcomes over free-software moralism - opposing strict GPL enforcement and tolerating lock-in - and positions the LLM transition as a structural shift toward corporate control at the expense of broad community participation.