AI coding is not the assistant on the side anymore — it's becoming the backbone of modern software workflows. Cursor’s 18-month Developer Habits Report cuts through the hype with the first hard evidence: AI isn’t just making engineers faster, it’s changing the very shape of software work. From autocomplete to system-level understanding, and from minor PR edits to letting AI manage “mega” pull requests, the ground is shifting under every developer. But Cursor’s report, based on their usage telemetry, also proves what most coverage misses: AI is widening the gap between developers, not closing it. For builders and engineering leads, this isn’t more “vibe coding” discourse — this is the data-backed call to rethink strategy, adoption, and training.
What does Cursor’s 18-month Developer Habits Report show?
Cursor’s Developer Habits Report is the first real measurement — not a survey, not anecdotes, actual usage data. The headline is clear: developer coding speed has roughly doubled year-over-year, and that pace is still climbing in 2026. Writers aren’t just churning out more small changes; the average lines of code added per pull request are up about 2.5× since last year, a sign that engineers are using AI to take on larger chunks of a project at once.
