--- title: Medium | daily.dev description: Three studies published in July 2026 — from CMU, Stanford, and an open-source analysis — converge on a troubling pattern: AI-generated pull requests receive... canonical: https://daily.dev/posts/medium-tkvoy3wa1 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Medium | daily.dev og:description: Three studies published in July 2026 — from CMU, Stanford, and an open-source analysis — converge on a troubling pattern: AI-generated pull requests receive... og:url: https://daily.dev/posts/medium-tkvoy3wa1 og:image: https://api.daily.dev/og/posts/tkvOY3Wa1.png og:image:alt: Medium og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Medium **[Medium](https://daily.dev/sources/medium_js)** · 7 min read · 0 upvotes · 0 comments ## Summary Three studies published in July 2026 — from CMU, Stanford, and an open-source analysis — converge on a troubling pattern: AI-generated pull requests receive less scrutiny, faster merges, and less discussion than human-written code, despite arguably needing more review. While raw code quality hasn't measurably dropped, review discipline is eroding as AI-driven volume rises. The post argues this risk is amplified in outsourced or staff augmentation contexts, where clients have limited visibility into a vendor's internal review culture. It offers four concrete questions to ask vendors about how they handle AI-generated code review, covering risk-based routing, decomposition of large diffs, measurement vs. felt velocity, and ownership of high-consequence merge decisions. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Do AI-generated pull requests get reviewed less than human-written pull requests? Yes. A Carnegie Mellon study analyzing 3,100 practitioner discussions found that agent-authored pull requests receive lower review rates, faster merges, and less discussion than human-authored ones. This is counterintuitive since AI-generated code arguably needs more scrutiny, not less, but the discourse data shows the opposite pattern playing out in practice. _Teams weighing AI coding workflows can track emerging code review research like this through daily.dev._ ### Does higher AI coding tool adoption reduce measured code quality in open-source projects? Not clearly. A study covering 13,360 AI chat sessions across 1,356 open-source repositories found no broad measured drop in code quality tied to AI agent adoption. The bigger risk identified is process-related: review discipline relaxing as output volume rises, rather than the code itself becoming worse. Developers also perceive others' AI-written code as harder to maintain than the data actually supports. _Anyone deciding how much to trust AI-assisted contributions can follow this kind of quality research on daily.dev._ ### What should a company ask an outsourced or staff augmentation vendor about how they review AI-generated code? Ask how they route review by risk rather than queue order, whether large AI-generated changes are decomposed into smaller reviewable pieces before review, what they measure to confirm quality instead of relying on how fast work feels, and who specifically owns the merge decision on high-consequence changes. Vague answers to these questions signal the vendor hasn't thought through AI-era review discipline. _Teams drafting vendor contracts for AI-assisted development can use daily.dev to stay ahead of review-process gaps like these._ ## Similar posts on daily.dev - [Agentic Code Review](https://daily.dev/posts/agentic-code-review-0ue8fcsa1) · Addy Osmani · 146 upvotes · 14 comments --- Tags: [#ai-coding](https://daily.dev/tags/ai-coding), [#code-review](https://daily.dev/tags/code-review) [View this post on daily.dev](https://daily.dev/posts/medium-tkvoy3wa1) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Medium","url":"https://daily.dev/posts/medium-tkvoy3wa1","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/medium-tkvoy3wa1"},"datePublished":"2026-08-01T04:59:42.022Z","dateModified":"2026-09-13T18:29:33.983Z","description":"Three studies published in July 2026 — from CMU, Stanford, and an open-source analysis — converge on a troubling pattern: AI-generated pull requests receive...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/87109fe46fed8715b9c0758a117f7edc?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/87109fe46fed8715b9c0758a117f7edc?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Medium","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Medium","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/medium","url":"https://daily.dev/sources/medium_js"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/medium-tkvoy3wa1","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-coding,code-review","timeRequired":"PT7M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Medium","item":"https://daily.dev/sources/medium_js"},{"@type":"ListItem","position":3,"name":"Medium"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/medium-tkvoy3wa1#faq","mainEntity":[{"@type":"Question","name":"Do AI-generated pull requests get reviewed less than human-written pull requests?","acceptedAnswer":{"@type":"Answer","text":"Yes. A Carnegie Mellon study analyzing 3,100 practitioner discussions found that agent-authored pull requests receive lower review rates, faster merges, and less discussion than human-authored ones. This is counterintuitive since AI-generated code arguably needs more scrutiny, not less, but the discourse data shows the opposite pattern playing out in practice. Teams weighing AI coding workflows can track emerging code review research like this through daily.dev."}},{"@type":"Question","name":"Does higher AI coding tool adoption reduce measured code quality in open-source projects?","acceptedAnswer":{"@type":"Answer","text":"Not clearly. A study covering 13,360 AI chat sessions across 1,356 open-source repositories found no broad measured drop in code quality tied to AI agent adoption. The bigger risk identified is process-related: review discipline relaxing as output volume rises, rather than the code itself becoming worse. Developers also perceive others' AI-written code as harder to maintain than the data actually supports. Anyone deciding how much to trust AI-assisted contributions can follow this kind of quality research on daily.dev."}},{"@type":"Question","name":"What should a company ask an outsourced or staff augmentation vendor about how they review AI-generated code?","acceptedAnswer":{"@type":"Answer","text":"Ask how they route review by risk rather than queue order, whether large AI-generated changes are decomposed into smaller reviewable pieces before review, what they measure to confirm quality instead of relying on how fast work feels, and who specifically owns the merge decision on high-consequence changes. Vague answers to these questions signal the vendor hasn't thought through AI-era review discipline. Teams drafting vendor contracts for AI-assisted development can use daily.dev to stay ahead of review-process gaps like these."}}]} ```