🔍 Read the full analysis: From Writing Code To Reviewing It: AI Agents Shift GitHub’s PR Bottleneck on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
GitHub’s Octoverse report says more than 40% of pull requests on its platform in September 2026 began with a detectable coding agent, up from 17% in March. The company has introduced AI review and model-routing tools as code review takes on a larger role, while saying developers remain responsible for changes that ship.
More than 40% of pull requests on GitHub in September began with a detectable coding agent, up from 17% in March, according to the platform’s Octoverse 2026 report, released October 28. The shift puts more pressure on code review: GitHub introduced new AI review and routing tools this week, but says developers remain accountable for code that ships.
GitHub says the September figure represents 110 million agent-started pull requests in one month, about six times the March volume. Its definition is specific: a pull request counts when its first draft came from a coding agent GitHub can detect. The company describes the observed figure as a floor because some agent use may go undetected. The report’s methodology includes public and private repositories unless a metric is marked otherwise; GitHub did not provide a public-versus-private split for this measure.
Many of those proposed changes were merged. GitHub reports that 90% of agent pull requests were merged within two weeks in September, compared with 83% in February. The rate for pull requests without a detected agent was about 86%. A merge records that code entered a repository; it does not establish that the change was correct, safe or free of later rework.
GitHub’s new offerings center on review and approval. Copilot code review Max combines reviews from several AI model providers into one result, while Copilot Fusion routes tasks to models for different steps. GitHub also made Copilot approvals eligible to count toward required approvals. Its estimated review prices are $0.25 for Lite, $1.50 for Balanced and $10 for 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 Becomes the New Constraint
The reported growth changes where engineering teams may spend time. If agents produce more first drafts quickly, people still need to inspect changes, test them and decide whether to approve them. GitHub’s launches reflect that pressure: the company is adding review capacity and tools to direct work among models, rather than focusing only on generating more code.
That shift has practical costs and governance consequences. Organizations may need to plan for review credits and computing use, set rules for where AI-generated approvals count, and track outcomes such as rework, reversions and incidents. Pull-request volume or merge rates alone cannot show whether software quality improved.
For developers, the report points to a changing mix of work: writing clear task instructions and critically examining a diff may matter more as agents take on routine implementation. The source argues that small starter tasks can be among the work agents absorb first, but it provides no measured data on effects for junior hiring or training.
Products Worth Considering
Logitech MX Mechanical Full-Size Wireless Mechanical Keyboard, Tactile Quiet Switches, Backlit Keys, Bluetooth, USB-C, Multi-Device, Graphit
Logitech MX Mechanical Wireless Illuminated Keyboard Tactile (Graphite)
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
As an affiliate, we earn on qualifying purchases.
How GitHub Counts Agent Work
Octoverse is GitHub’s annual report on activity across its platform. The 2026 edition covers October 1, 2025, through September 30, 2026. GitHub’s reported agent share uses detectable signals and defines the counted pull request by who produced its first draft. A human may have substantially revised the code afterward and the pull request can still count as agent-started.
The report uses several related labels, including “agent-started,” “agent-generated” and “agent-authored.” For the headline statistic, agent-started is the more precise description because it matches the definition based on the first draft. GitHub also recalculated figures for 2024 and 2025 using updated definitions and time windows, so comparisons with older Octoverse editions may not be direct.
The agent trend sits within broader platform growth. GitHub says merged pull requests reached 1.3 billion during the report year, up from 538 million in 2025 and 402 million in 2024. Those totals reflect all pull requests, not just agent-started ones, and do not by themselves identify the causes of the increase.
“A pull request whose first draft came from a coding agent that we can detect.”
— GitHub, in the Octoverse 2026 report
What Merge Rates Cannot Show
The report does not establish whether agent-started pull requests are as reliable or maintainable as other changes. The 90% two-week merge rate measures acceptance into a repository, not defect rates, security outcomes, rework or long-term performance. GitHub’s figures also omit agent activity it cannot detect, and the company has not published a split showing how the 40%-plus share differs between public and private repositories.
The precise financial effect of the new review tools on a team is also unsettled. GitHub gives estimated per-review prices and notes additional Actions minutes, but the source does not specify how many reviews a typical pull request will require or provide a representative total cost. Nor does the report quantify how the change affects staffing, junior developer opportunities or the amount of human review needed.
Teams Must Test Review Outcomes
GitHub has released the review and routing features, and teams can decide whether to adopt them and whether AI approvals should count toward their own required checks. The next useful evidence will be operational: review time, defect and incident rates, rework, reversions and total costs, tracked alongside the number of agent-started pull requests and the conditions under which they were approved.
GitHub’s current report provides a platform-wide snapshot through September 2026, not a forecast of future adoption or a verdict on code quality. Further reporting will be needed to determine whether AI review tools reduce the human review burden, shift it to different tasks or add another layer of cost. For now, the confirmed change is the scale of agent participation and GitHub’s response with tools aimed at checking and routing the resulting work.
Key Questions
What does GitHub mean by an agent-started pull request?
GitHub counts a pull request when its first draft came from a coding agent the company can detect. Human revisions afterward do not necessarily change that classification.
Did agents start more than 40% of pull requests in every month?
No. The report says the share was more than 40% in September 2026, compared with 17% in March. It does not say that the September share applied to every month.
Does a high merge rate prove agent-written code is good?
No. GitHub says 90% of agent pull requests were merged within two weeks in September, but merging is not a measure of correctness or long-term quality. The report does not provide comparable defect or rework rates.
How much does GitHub’s AI code review cost?
GitHub estimates an average review at $0.25 for Lite, $1.50 for Balanced and $10 for Max. GitHub Actions minutes are additional, and actual costs will depend on use.
Who is responsible when an AI-generated change ships?
GitHub says developers remain accountable. It also says its cloud agent cannot approve or merge its own pull request, and the person who requested the change cannot approve it.
Source: ThorstenMeyerAI.com
