Microsoft wants to remove all C and C++ code from the company by the end of the decade. Distinguished Engineer Galen Hunt proposes an ambitious plan in which AI and algorithms will rewrite code bases en masse to Rust. The goal at least sounds like a decent slogan: one engineer, one month, one million lines of code.
Microsoft wants to get rid of technical debt with the help of modern tooling. The company has decades of old programming code, mostly written in the old language C and one of its many derivative languages, C++. The desire to replace this code is also a desire to eliminate an entire class of vulnerabilities: memory-related bugs. Authorities around the world have been saying for some time that tech companies need to fill in their roadmaps to emphasize memory safety. This includes the use of the low-level, memory-safe language par excellence: Rust.
Microsoft is therefore committed to eliminating this technical debt. It is attempting to do so through modern tooling. The company has developed a powerful infrastructure for code processing that combines AI agents with algorithmic analysis. That infrastructure is already running at scale for problems such as code comprehension, Hunt writes on LinkedIn.
Urgency for memory-safe code
This step did not come out of the blue, and is not based solely on warnings from external cyber experts. In 2023, Microsoft announced that it would rewrite parts of the Windows kernel in Rust after Azure CTO Mark Russinovich banned developers from starting new C/C++ projects. Microsoft was “all-in on Rust,” Russinovich said earlier this year. The company has since dramatically expanded its use of the memory-safe programming language.
The motivation behind this change of course is clear: memory errors in C and C++ cause an estimated 70 percent of all vulnerabilities in Microsoft products. Rust prevents these problems by building in memory safety without sacrificing performance. Those who circumvent certain functionality can still create memory leaks with Rust, but it is much less obvious than with C and C++. Those who use Rust as recommended should be immune to memory safety issues.
Russinovich previously reported that Microsoft was working on “more automated translation from C and C++ to Rust using LLMs.” Hunt’s LinkedIn post is the concrete follow-up to that. He is now looking for a Principal Software Engineer to realize this ambition.
Job opening requires Rust expertise
The job requirements are as follows: candidates should preferably have at least three years of experience writing systems-level code in Rust. Compiler, database, or OS implementation experience is highly desirable. Although compiler experience is not mandatory, the selected candidate must be willing to build that expertise within the team.
Hunt’s Rust refactoring team is part of the Future of Scalable Software Engineering group within Microsoft CoreAI’s Engineering Horizons organization. Its mission is to “build capabilities that enable Microsoft and our customers to eliminate technical debt at scale.” The team works with internal customers and partners on new tools and techniques, which are then rolled out across the company and industry.
Rust adoption is growing worldwide
Microsoft is part of a broader enthusiasm for using Rust. According to JetBrains, 2.3 million developers use the programming language, with 709,000 considering it their primary language. In enterprise environments, adoption increased by 68.75 percent. In system programming, Rust’s share grew from 1.05 percent in 2024 to 1.47 percent in 2025, a relative increase of 40 percent. Rust is a relatively complex language compared to Python, for example, and does not have the decades of expertise of programmers as is the case with C and C++, but as a low-level language, it should not have too many layers of abstraction. The idea behind the programming language is to run critical systems such as kernels, IoT devices, and system drivers.
Google, Amazon, Dropbox, and Cloudflare now make extensive use of Rust for critical infrastructure. Google even invested $1 million in the Rust Foundation to improve interoperability between C++ and Rust. This relationship between new and old also leads to friction in the development of the Linux kernel, partly because old hands in its development are not used to reviewing code other than C and C++. The Rust project itself also has its challenges, to say the least. A stable future for the future core of IT infrastructure is getting ever-more important as tech giants adopt it.