🔍 Read the full analysis: Why AI-Started GitHub Pull Requests Put Review Processes In The Spotlight on ThorstenMeyerAI.com
Get everyday 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 increase is drawing attention to review capacity and approval safeguards; GitHub says agent-written changes still require human accountability, while its new review options carry per-review costs.
More than 40% of GitHub pull requests in September 2026 began with a detectable coding agent, up from 17% in March, according to the platform’s Octoverse report released October 28. The shift—about 110 million agent-started pull requests in one month, GitHub says—puts pressure on teams to review changes at a pace that matches the growing volume of code.
GitHub defines an agent-started pull request as one whose first draft came from a coding agent the company can detect. Its reported figure is a floor, the company says, because some agent use may go undetected. The measure includes public and private repositories unless otherwise noted; GitHub did not provide a public-versus-private breakdown for this figure.
GitHub also reported that 90% of agent pull requests were merged within two weeks in September, compared with 83% in February. The merge rate for pull requests without a detected agent remained about 86%. A merge records that a change was accepted into a codebase; it does not, by itself, show that the code was correct or free of later problems.
The company’s announcements at its Universe conference focused on review and routing as well as code generation. GitHub introduced Copilot code review Max, which combines reviews from several AI models, made Copilot approvals eligible to count toward required approvals, and released Copilot Fusion, a system that routes different tasks to models. GitHub estimates that an average Copilot review costs $0.25 in Lite, $1.50 in Balanced and $10 in Max, in addition to GitHub Actions minutes.
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 Faces a New Test
The figures point to a change in where software teams may spend time. If agents draft more pull requests, human review can become the constraint: reviewers still need to understand the proposed change, check its effects and decide whether it is safe to merge. That work does not disappear simply because a first draft arrives faster.
For organizations, the report makes review a question of staffing, workflow and cost, not only model access. GitHub’s per-review estimates offer a way to think about direct charges, but teams also need to account for reviewer time and any work required after a merge. Pull-request volume or merge rate alone cannot establish whether the system is improving software quality.
For developers, the shift could change the value of everyday skills. Writing a clear task for an agent and inspecting a multi-file diff critically both become more consequential when a tool produces the initial change. The source report notes that 87% of agent pull requests changed more than one file, compared with 79% of pull requests without a detected agent. That comparison indicates broader changes are common in the agent group, though it does not show whether they are harder to review or more error-prone.
Products Worth Considering
Amazon Kindle Paperwhite (12th Generation, 2024), 16 GB, 7-inch Display, Jade
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Black
Apple iPad Air 11-inch (M2), 512GB, Wi-Fi, Space Gray
As an affiliate, we earn on qualifying purchases.
How GitHub Defines Agent-Started
The numbers come from Octoverse 2026, GitHub’s annual report on activity across its platform. This edition covers October 1, 2025, through September 30, 2026. GitHub says the agent share rose from 17% in March to more than 40% in September, and describes September’s total as 110 million pull requests. The source says that volume was about six times March’s, but the reported share and total use different measures: one is a proportion, the other a count.
The wording matters. GitHub’s definition counts a pull request when its first draft came from a detectable agent, even if a person later rewrote part of it. The report uses other labels in charts and industry commentary, including “agent-generated,” “agent-authored” and “agent-opened.” Its glossary definition is the clearest basis for interpreting the headline figure.
GitHub recalculated its 2024 and 2025 figures under updated definitions and reporting windows, so older Octoverse numbers are not directly comparable. The wider platform also saw substantial growth: the report says merged pull requests rose from 402 million in 2024 and 538 million in 2025 to 1.3 billion in the year covered. That broader rise is relevant when considering review workloads, but it does not isolate the effect of agents.
“A pull request whose first draft came from a coding agent that we can detect.”
— GitHub’s Octoverse report
What the Merge Figures Cannot Show
The report does not establish whether agent-started pull requests have the same defect rates, security outcomes or long-term maintenance costs as other changes. A high merge rate is not a quality measure, and GitHub’s published figures do not provide a comparison of bugs, reverts, incidents or rework for the two groups.
It is also unclear how much agent use the detection method misses, how the reported activity divides between public and private repositories, and how much human editing occurs after an agent’s first draft. Because GitHub describes the count as a floor, the true share could be higher, but the source does not quantify the gap. Nor does the report show how many teams will use AI review or how often its approvals will count toward required review thresholds.
Teams Set Review and Approval Rules
GitHub’s newly announced review features give teams additional options, but organizations will need to decide where AI review fits in their approval policies and how much review spending to allow. The source material does not specify a universal rollout timetable or say that every repository will adopt these settings.
For teams adopting agents, the next practical test is whether faster drafting produces better outcomes after review and release. That means tracking indicators such as rework, reverted changes and production incidents alongside pull-request counts and merge times. GitHub’s report supplies platform-wide activity figures; it does not settle how those outcomes compare across teams or projects.
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 it can detect. A person may edit the draft afterward and it can still count.
Does a 90% merge rate mean agent-written code is reliable?
No. The figure means 90% of agent pull requests were merged within two weeks in September, according to GitHub. It does not measure correctness, defects or later maintenance needs.
How much does Copilot code review cost?
GitHub estimates an average review at $0.25 for Lite, $1.50 for Balanced and $10 for Max, plus GitHub Actions minutes. These are the company’s estimates, not a measure of total reviewer or follow-up costs.
Who is responsible for code an agent helps produce?
GitHub says developers remain accountable for code that ships. Its source material also says its cloud agent cannot approve or merge its own pull requests, and the person who requested a change cannot approve it either.
Source: ThorstenMeyerAI.com
