Kotlin at 15: JetBrains finds 93% of developers using AI

Kotlin has turned fifteen, and JetBrains has used the anniversary to publish a report that tries to answer a blunter question than the milestone suggests: does the language still have somewhere to go?

The State of Kotlin in 2026 report, published on the JetBrains Kotlin blog, blends the company's own research with independent industry data and real-world adoption evidence.

The AI numbers are the headline

Among surveyed Kotlin developers, 93% use at least one AI coding tool, and 81% already use AI coding agents or say they are likely to try them.

Confidence in the output is high. An identical 81% say AI-generated Kotlin code comes back either fully correct or needing only minor fixes.

JetBrains argues that Kotlin's existing design choices - static typing, null safety and coroutines - are what make it a reasonable fit for AI-assisted work rather than a language being dragged along by it.

It is no longer just an Android language

The report's second thread is about drift away from mobile. More than half of Kotlin developers now report working on backend projects, and JetBrains frames that as the clearest evidence the language has outgrown its original home.

Multiplatform work features alongside backend and AI-assisted development in the report's framing, and the company says the community itself has grown older and more experienced, with fifteen years now long enough to separate the trends that lasted from the ones that did not.

JetBrains says the findings partly confirmed expectations and partly challenged assumptions it held. The full report and a downloadable PDF are published alongside the blog post.

Our opinion

Vendor research always deserves a sceptical read, and this one is not neutral: JetBrains sells the IDE, the AI assistant and the language, so a report concluding that Kotlin suits the AI era is a report concluding that its own product line does.

That said, the 93% figure is worth sitting with, because it is not really a Kotlin statistic. It is a measurement of how completely AI tooling has moved from experiment to default across professional development, and the number is only surprising if you were still treating coding assistants as an opt-in curiosity.

The more interesting claim is the backend drift. Kotlin's reputation was built almost entirely on Android, and a majority of its developers reporting backend work is the kind of quiet migration that matters more to a language's next fifteen years than any anniversary post.

The honest caveat is that self-reported adoption surveys reward enthusiasm, and 81% finding AI output acceptable says as much about how developers grade generated code as it does about the models. Read the report for direction of travel, not for precise numbers.