🔍 Read the full analysis: Can GitHub Teams Review Pull Requests As Fast As AI Agents Start Them? on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- 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 the platform in September 2026 began with a detectable coding agent, up from 17% in March. The company introduced AI review and model-routing tools as human review becomes a larger part of the development process; GitHub says developers remain accountable for code that ships.
More than 40% of pull requests on GitHub in September 2026 began with a detectable coding agent, according to the platform’s annual Octoverse report, up from 17% in March. The report’s release at GitHub Universe on October 28 came alongside new AI review and model-routing tools, as the company focuses on how developers check and approve the growing volume of agent-written code.
GitHub said agents started 110 million pull requests in September. It defines an agent-started pull request as one whose first draft came from a coding agent the company can detect. GitHub describes the count as a floor because some agent use may not be detected. The figure includes public and private repositories under the report’s general methodology, but GitHub did not provide a split between them.
The report says 90% of detected agent pull requests were merged within two weeks in September, compared with 83% in February. Pull requests without a detected agent held at about 86%. A merge records that a change was accepted into a codebase; it does not establish that the code was correct, safe or free of later rework.
GitHub announced Copilot code review Max, which uses several AI reviewers from different model providers and combines their feedback. The company also said Copilot review approvals can count toward required approvals, and introduced Copilot Fusion, a system that routes tasks to models. GitHub estimates average review costs of $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 Capacity Becomes the Constraint
The reported growth shifts attention from how quickly code can be drafted to whether teams can inspect and validate changes at the same pace. A pull request is a proposed change that ordinarily needs review before it enters a product. If agents produce more proposals, teams may need additional reviewer time, clearer routing and more deliberate limits on when automated feedback counts toward approval.
The figures also point to a cost that can be easy to miss when measuring AI adoption. GitHub’s estimates put a price on each Copilot review tier, while Actions minutes add another expense. For companies, those costs sit alongside the human time needed to judge the change and its tests. Counting agent-started or merged pull requests alone does not show whether a workflow saves time or improves software quality.
GitHub says developers remain accountable for what ships. Its cloud agent cannot approve or merge its own pull requests, and the person who requested a change cannot approve it either. Those controls leave a human sign-off in the process, but they do not by themselves establish how thoroughly a reviewer examined a patch.
Popular right now
Dyson Cyclone V10 Origin Cordless Vacuum
Shark Navigator Robot Vacuum with 30-Day Self-Emptying Base, White (AV2110S)
KitchenAid Artisan Series 5-Quart Tilt-Head Stand Mixer, Dried Rose (KSM150PS)
As an affiliate, we earn on qualifying purchases.
How GitHub Defines Agent Work
The Octoverse 2026 report covers October 1, 2025, through September 30, 2026. GitHub’s central measure is narrower than a count of every pull request touched by AI: it covers a pull request when the first draft came from an agent the platform can detect. A human can substantially rewrite that draft and the pull request still counts as agent-started.
The report uses several nearby terms, including “agent-generated,” “agent-authored” and “agent-opened.” Its definition makes “agent-started” the most precise wording for the headline share. GitHub also recalculated its 2024 and 2025 figures under updated definitions and time windows, so comparisons with earlier Octoverse reports are not direct.
The agent increase comes amid broader platform growth. GitHub reported 1.3 billion merged pull requests during the year to September, up from 538 million in 2025 and 402 million in 2024. It also said 55% of Copilot’s monthly active users used agent features, compared with 35% a year earlier. Those are platform-reported measures, not independent assessments of code quality or productivity.
“a pull request whose first draft came from a coding agent that we can detect”
— GitHub, in the Octoverse 2026 report
What the Metrics Cannot Show
The report does not say how much of the 40%-plus share came from public repositories versus private company code, nor how many agents GitHub could not detect. Its definition also counts a pull request based on the origin of its first draft, not on how much of the final code an agent wrote. The total therefore should not be read as a precise measure of AI’s contribution to finished software.
GitHub’s merge-rate figures do not show whether merged changes later needed substantial rework, were reverted or caused incidents. The source material provides no comparable measures of defects, review quality or time spent by human reviewers. It is also unclear how often teams will use Copilot approvals to meet repository requirements and how that will affect review practices across different projects.
The stated review prices are estimates for the named tiers, not a full cost of operating an AI-assisted development workflow. Actions minutes are additional, and the report does not provide a typical monthly bill or an independent comparison of the tool’s results against human-only review.
Teams Must Track Review Outcomes
GitHub has released the review and routing features, but the report does not provide a date by which every organization must adopt them. Teams will need to decide which repositories permit Copilot approvals to count, which review tier fits their needs and how to allocate human reviewers as agent activity grows.
The more informative test will be what happens after a pull request is merged. Organizations can compare review time, rework, reversions and production incidents alongside usage and spending to assess whether agent-assisted workflows help. GitHub’s September numbers describe adoption and merging; they do not settle whether teams can review the resulting code as quickly, or whether the code performs well after release.
Key Questions
What does GitHub mean by an agent-started pull request?
It is a pull request whose first draft came from a coding agent GitHub can detect. A human may later rewrite part of it and it still qualifies under that definition.
How many pull requests did agents start in September?
GitHub reported 110 million, representing more than 40% of pull requests on the platform that month. The company says its detection measure is a floor because some agent use may go uncounted.
Does a high merge rate show that agent-written code is good?
No. GitHub reported that 90% of detected agent pull requests were merged within two weeks, but merging does not establish correctness or quality. The report does not give comparable figures for later rework, reversions or incidents.
How much does GitHub estimate Copilot code review costs?
GitHub estimates an average review at $0.25 for Lite, $1.50 for Balanced and $10 for Max. GitHub Actions minutes are additional; these estimates are not a complete measure of total workflow costs.
Who approves and takes responsibility for agent-written changes?
GitHub says developers remain accountable for shipped code. Its cloud agent cannot approve or merge its own pull request, and the person who requested a change cannot approve it either. Teams still need to set and follow their own review practices.
Source: ThorstenMeyerAI.com
