AI Agents Open 40 Percent of GitHub Pull Requests
Autonomous tools now represent a major portion of coding activity as of September 2026.
Updated on Oct. 28, 2026 in Artificial Intelligence
Live Poll
Do you trust AI agents to handle complex professional tasks without human review?
AI agents opened more than 40% of all pull requests on GitHub in September 2026, totaling 110 million requests. This reflects a rapid increase from the 17% share recorded in March 2026.
Why it matters
The shift highlights the growing reliance on cloud-based autonomous agents that perform tasks remotely rather than on local developer machines. These agents now play a central role in managing code workflow across the platform.
AI-driven pull requests account for 40% of activity, with 90% of these requests merging within two weeks. Additionally, 55% of monthly active GitHub Copilot users now leverage agentic features for their daily tasks.
The players
GitHub
This is a global software development platform that hosts billions of lines of code and provides tools for collaborative version control.
Satya Nadella
He is the CEO of Microsoft, which owns GitHub, and is a prominent leader in the global technology and artificial intelligence sectors.
The details
Repositories created in 2026 now include an AI agent configuration file that provides specific instructions on project operations. Agent-involved pull requests show a distinct pattern, as they change multiple files 87% of the time compared to 79% for traditional pull requests.
Timeline
March 2026: AI agents accounted for 17% of total pull requests.
September 2026: AI agents opened 40% of total pull requests.
September 30, 2026: The annual reporting period for GitHub data concluded.
October 28, 2026: Satya Nadella presented the Octoverse 2026 report findings.
The Tech Race
The rise of autonomous coding agents follows the widespread industry integration of GitHub Copilot. This marks a fundamental evolution in software engineering as platforms transition from manual code writing to managing AI-led autonomous workflows.
Developers are increasingly offloading repetitive tasks to cloud-based agents, which requires adopting new configuration files for their repositories. Users can expect faster code integration, as agent-driven requests currently maintain a 90% merge rate.
The takeaway
As autonomous coding becomes the new standard, developers should focus on mastering agent configuration to maintain control over project operations. Understanding how these tools handle multi-file changes is essential for maintaining code quality in an AI-assisted environment.
Further reading
Learn more about the latest industry trends in Artificial Intelligence.
Live Poll
Do you trust AI agents to handle complex professional tasks without human review?
