🔍 Read the full analysis: AI Agents Move GitHub Pull Requests Forward, While Review Takes More Focus on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
GitHub’s Octoverse report says more than 40% of pull requests on its platform began with a coding agent’s first draft in September 2026, up from 17% in March. The company introduced tools for AI-assisted review and model routing, while saying developers remain accountable for code that ships. Its counts are based on agents GitHub can detect, and a merged pull request is not proof that code is correct.
More than 40% of GitHub pull requests began with a coding agent’s first draft in September, up from 17% in March, according to the company’s Octoverse report released October 28. The rise is shifting attention toward how developers review and approve code: GitHub announced new AI review tools and approval controls, while saying people remain accountable for changes that ship.
GitHub counted 110 million agent-started pull requests in September, about six times the March volume, according to its report. It says 90% of detected agent pull requests were merged within two weeks in September, compared with 83% in February. For pull requests without a detected agent, the reported merge rate stayed around 86%. Those figures track whether a change was merged, not whether it was correct, safe or free of later rework.
The company defines an agent-started pull request as one whose first draft came from a coding agent GitHub can detect. It describes the count as a floor because some agent use may go undetected. The report’s methodology includes public and private repositories unless otherwise specified, but GitHub did not provide a public-versus-private breakdown for this measure. Its wording varies in places; “agent-started” is the specific definition behind the headline figure.
GitHub’s new offerings focus on review and workflow. Copilot code review Max uses reviewers from different model providers and combines their feedback. GitHub also made Copilot approvals eligible to count toward required approvals and released Copilot Fusion, a system that routes tasks to models. The company estimates an average review will cost $0.25 in Lite, $1.50 in Balanced or $10 in Max, with GitHub Actions minutes charged separately.
Infographic 01 · The agent share
Four in ten pull requests now start with an agent.
GitHub’s own telemetry, March to September 2026. The first draft comes from an agent; review and merge stay with people.
Merged within two weeks
Share of pull requests merged within 14 days
Merged fast is not the same as correct. GitHub publishes no revert or defect rate.
Bigger changes to review
Share of pull requests that change more than one file
More files per change means more for a reviewer to check.
What GitHub counts: an “agent-started pull request” is one “whose first draft came from a coding agent that we can detect.” Undetected agents are missing, so GitHub calls the figure a floor. Metrics cover public and private repositories unless noted.
Source: GitHub Octoverse 2026, 28 Oct 2026 (text, chart captions, glossary and methodology). All figures are GitHub’s own.
Powered by Thorsten Meyer AI
Review Work Is Becoming the Constraint
When agents produce more first drafts, the amount of code awaiting human judgment can rise even if writing each change takes less time. The key operational question is no longer just how quickly a team can produce a pull request, but whether reviewers can check its logic, scope and effects before it is merged. GitHub reports that 87% of agent-started pull requests changed more than one file, compared with 79% of pull requests without a detected agent. A broader change can require reviewers to track more interactions across a codebase.
The new tools could help teams distribute review work, but they also introduce per-review charges and questions about when an AI-generated approval should count. Organizations will need to set policies for which changes require a human sign-off and track outcomes such as rework, reversions and incidents, rather than treating a high merge rate or pull-request count as evidence of quality. For developers, writing a clear task and critically reading a diff become more central as agents take on more drafting.
GitHub’s report also points to a workforce concern: small, bounded coding tasks have often given newer developers practice. If agents increasingly handle those tasks, teams may need to make deliberate room for junior developers to learn through review, maintenance and supervised changes. The report documents adoption, but does not establish how this shift affects hiring, productivity or software quality.
Products Worth Considering
Logitech MX Mechanical Wireless Illuminated Keyboard Tactile (Graphite)
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
Logitech MX Keys S Wireless Keyboard, Graphite
As an affiliate, we earn on qualifying purchases.
How GitHub Counts Agent Activity
The Octoverse report is GitHub’s annual review of activity on its platform. Its 2026 edition covers October 1, 2025, through September 30, 2026. GitHub recalculated figures for 2024 and 2025 using updated definitions and reporting windows, so comparisons with older Octoverse editions may not be direct.
Activity expanded across the platform during the reporting period. GitHub says developers made 17.9 billion pushes and 34.3 billion commits, while merged pull requests rose to 1.3 billion, from 538 million in 2025 and 402 million in 2024. These broader totals provide scale, but they do not isolate how much growth was caused by agents. GitHub also says that 55% of Copilot’s monthly active users used agent features, compared with 35% a year earlier; use of its cloud agent grew 230% year over year, according to the report.
The source material says GitHub’s telemetry showed activity rising around June, when the company described agents as moving from single-session tools toward persistent cloud-based work. That timing is GitHub’s account of platform activity, not independent evidence that the change in agent use alone drove the increase in pull requests.
“A pull request whose first draft came from a coding agent that we can detect.”
— GitHub, in its Octoverse report
Detection Does Not Measure Code Quality
The true share of agent-started pull requests is unknown because GitHub says its detection misses some agent use. The company has not published a breakdown showing how the September figure differs between public and private repositories, or how much code people rewrote after an agent supplied the first draft. GitHub’s report also uses related labels such as “agent-generated” and “agent-authored,” although its defined metric is specifically about how a pull request began.
The merge figures do not establish that agent-written code is as reliable as human-written code. The source material provides no matched analysis of defects, security issues, rework or production incidents across the two groups. It also does not say how often AI review comments catch problems that human reviewers would miss, or whether AI approvals affect the quality of merged changes. The listed review prices are GitHub estimates; actual team costs may vary with use and additional Actions minutes.
Teams Must Set Review Policies
GitHub has released its report and announced the review and routing features, but the source material does not give a timetable for further rollout milestones or independent evaluations. Teams considering the products will need to decide which review tier fits their work, how to budget for review credits and Actions minutes, and whether Copilot approvals can satisfy particular repository rules.
The next useful evidence will come from how organizations measure the results: review time, defects, rework, reversions and incidents, alongside pull-request volume and merge rates. GitHub’s figures establish a fast rise in detected agent-started drafts; they do not resolve whether review tools can keep pace while maintaining code quality. Developers and engineering managers will continue to decide what is approved and what is shipped under the accountability rules GitHub describes.
Key Questions
What does GitHub mean by an agent-started pull request?
It means a pull request whose first draft came from a coding agent GitHub can detect. A person may later revise the draft and it can still count under that definition.
How many GitHub pull requests started with an agent in September?
GitHub reported 110 million agent-started pull requests in September 2026, representing more than 40% of its pull requests for that month. The company says its detection misses some agent use, so the figure is a floor.
Does a high merge rate mean agent-written code is correct?
No. A merge rate shows that a pull request was merged within a specified period; it does not establish that the code was correct or free of later problems. GitHub’s report does not provide a quality comparison based on defects or rework.
What do GitHub’s new AI review options cost?
GitHub estimates an average Copilot review at $0.25 for Lite, $1.50 for Balanced and $10 for Max. GitHub Actions minutes are additional, and the source material does not give a single all-in cost for every team.
Who is responsible for code an AI agent proposes?
GitHub says developers remain accountable for code that ships. It also says its cloud agent cannot approve or merge its own pull requests, and the person who requested the change cannot approve it either.
Source: ThorstenMeyerAI.com
