Canonical is co-funding a three-year PhD project at the University of Bristol. The research investigates whether artificial intelligence can translate large C codebases into safe Rust. Engineering vice president Jon Seager announced the investment on Ubuntu's Discourse forum.
The initiative does not mean Canonical is turning bots loose to rewrite Ubuntu. Instead, the company is paying a researcher to spend several years testing whether the idea works. This approach provides empirical evidence in an industry full of hype.
Under Seager's guidance, Ubuntu 25.10 adopted a Rust implementation of sudo and the Rust-based uutils coreutils. However, those were pre-existing human-written replacements. The new project explores whether an LLM can decompose programs with hundreds of thousands of lines of C into smaller components and rewrite them in safe and maintainable Rust.
Traditional source-to-source translators often preserve C structure too literally. They rely heavily on unsafe operations and awkward idioms. Mature codebases also contain undocumented knowledge from years of fixes for real-world corner cases. Joel Spolsky highlighted this phenomenon in his 2000 essay.
The proposal examines two specific tools named snap-confine and AppArmor. Enterprise Linux has largely moved to SELinux, but Debian and other distributions still support AppArmor. A hardened Rust version could help AppArmor users.
UK Research and Innovation is co-sponsoring the project with Canonical. Seager will oversee the work alongside Professor Meng Wang and Dr Cristina David from the University of Bristol.
The research will run for three years, which is a standard duration for a UK PhD. More evidence is required to determine if generative AI can solve this translation problem.
