The 2025 Rust Survey dropped and the numbers are striking. Not because Rust is popular — we knew that. But because the adoption curve is bending in a way that suggests this isn't hype anymore. It's infrastructure.
The Headline Numbers
- 48.7% of companies now use Rust in production
- 68% of respondents said memory safety was a "primary motivation" for choosing Rust
- 42% reported their organization increased Rust investment in the past year
- Only 12% said they were "evaluating" Rust — the rest are either using it or not interested
This isn't a language finding its audience. This is an industry shifting its foundations.
What Drives Adoption
The survey breaks down motivation by sector:
| Sector | Primary Driver | |--------|---------------| | Web Services | Performance + safety | | Systems/Embedded | Memory safety | | Dev Tools | Correctness + tooling | | Finance | Reliability + regulatory pressure |
Notice what's absent from every category: "because it's trendy." The adoption is being driven by engineering necessity, not developer enthusiasm. That matters because it means it's sticky. Companies aren't trying Rust — they're betting on it.
The Enterprise Infiltration
The most significant shift isn't percentage points. It's who is using Rust now:
- Toyota: Production vehicles with Rust in safety-critical systems
- NASA: Formal verification pipelines with Rust backends
- FedRAMP: Memory safety mandates for cloud deployments
- Meta: 2.1M lines of Rust across infrastructure
- Microsoft: Azure SDKs, Windows subsystems, Edge
This isn't the Rust of 2015 — a language enthusiasts loved but no one bet their career on. This is the Rust of 2025, where your car, your cloud, and your government are all running code you can't afford to have buffer overflows in.
The Language Gap
Here's what concerns me: the survey also shows a widening gap between what developers want and what they deliver.
- 73% want async/await to be "stable and fast"
- 61% report friction with the borrow checker in async contexts
- 47% say they "work around" lifetimes rather than understand them
We're building critical infrastructure with a language that a majority of its users don't fully understand. This isn't a knock on Rust — it's a call to improve our tooling, our docs, and our teaching.
The borrow checker is a feature, not a hurdle. But 47% working around it suggests we haven't explained why well enough.
What Changes Next
Three trends to watch from this survey:
-
The C++ exodus accelerates. Companies with 10M+ lines of C++ are now actively migrating. Not in a single rewrite — but in new services, in side projects, in the places where memory safety matters most.
-
Rust becomes "invisible." Just like you don't think about TCP/IP when you build a web app, Rust will become the substrate. The language stops being the story; the systems it powers become the story.
-
The ecosystem matures past the language. The survey's "most needed improvements" aren't about syntax anymore. They're about: IDE support, build times, cross-compilation, and documentation. These are infrastructure problems, not language problems.
The Bottom Line
48.7% in production isn't a ceiling. It's a floor.
The memory safety conversation has moved from "if Rust" to "when Rust" to "how fast can we move." The survey confirms what the job market, the GitHub trends, and the enterprise announcements have been screaming: Rust isn't the future of systems programming.
It's the present.
What number surprised you most? Drop a note — I read everything.