Microsoft and Open AI build GitHub Copilot, an AI-based tool to recommend code to software developers in almost any programming language
- GitHub, GitHub's parent company Microsoft and OpenAI have teamed up to deliver a tool that comes up with source code for programmers to use as they work.
Context & Ripple Effects
This is the origin point of a four-year arc that the rest of the coverage fills in. Within months of launch, GitHub reported as much as 30% of new code on its network was being written with Copilot and that half of developers who tried it kept using it — early evidence the autocomplete-for-code idea stuck.
From there the product compounded: an in-depth look found Copilot writing 35% of its users' code, GitHub took it commercial at $10/month after a 12-month preview with 1.2M+ testers, then layered on GPT-4-powered chat and voice via Copilot X, before eventually shipping a coding agent and open-sourcing Copilot in VS Code in 2025. What started as a code recommender became the template for embedding AI directly inside developer tools.
First-order effects
- Developers get an assistant that suggests source code in-editor across most programming languages — GitHub, Microsoft, and OpenAI each gain a flagship proof point that OpenAI's models work as productivity software, not just demos.
Second-order effects
- Once adoption proved sticky enough to charge for, every tool vendor with an IDE or editor faced pressure to ship its own AI completion layer, and GitHub converted free preview usage into a $10/month subscription line.
Third-order effects
- If the pattern holds — recommend, then chat, then delegate — coding shifts from developers typing every line to developers reviewing AI-written code, with the editor itself becoming the battleground where model providers lock in workflows.
The trend: Developer tooling is absorbing generative AI from autocomplete toward agentic assistance, turning editors like VS Code into distribution chokepoints for AI models.