AI coding agents: 41% of teams say quality improved, 37% say it's faster but worse

The mabl 2026 State of Quality Engineering report finds teams using AI coding agents split almost evenly between better quality and faster but lower-quality code. Teams spend 20% of the week verifying AI-generated tests, and 35% of bugs are still found by customers.

Among teams using AI coding agents, opinion on quality is split almost down the middle. The mabl 2026 State of Quality Engineering report finds 41% say code quality has improved and 37% say the code is faster but lower quality. Same tools, opposite outcomes.

Why the split exists

The report points to what happens after the code is written. Teams spend on average 20% of their working week manually verifying AI-generated tests, and 35% of production bugs are still discovered by customers rather than by the team. Where checks are automated and owned, quality rises. Where AI output is simply merged because it looks fine, bugs reach users.

That reading matches other 2026 data. Sonar’s developer survey found AI now accounts for around 42% of committed code, yet only 48% of developers always check it before committing.

The practical takeaway for non-technical buyers

The tool is not the differentiator. Claude Code, Cursor and Copilot are all capable. The difference is the process wrapped around them: review, test coverage, and release discipline. Two agencies using the same AI tools can deliver very different products.

A useful test is to ask how bugs are found. If the honest answer is “users tell us”, you are buying the 37% outcome.

So what

For founders and product leads, “built with AI” should now prompt a follow-up question rather than reassurance: how is it verified? Choose a partner who can show you test coverage, review practice and a release process, not just a fast demo. BuildApps combines AI-assisted speed with that discipline across custom software development and mobile builds. If you are weighing a project, start a conversation.

← All research

More research.

GitHub Copilot adds Supabase, Webflow, Wix and five more connectors in public preview

GitHub Copilot now connects to eight more services, including Supabase, Webflow, Wix and Mixpanel, through verified MCP connectors in public preview. They join more than 25 existing ones.

Read ›

Gemini 4 Neon is generally available in GitHub Copilot, aimed at lower-cost agentic coding

Gemini 4 Neon is now generally available in GitHub Copilot for agentic coding. Google says its quality is close to Gemini 4 Carbon at a lower token cost per task.

Read ›

Claude Code 2.1.328 brings auto mode to Bedrock, Vertex and Foundry, and narrows /code-review to bugs

Claude Code 2.1.328 extends auto mode to Claude Opus 4.6 and Sonnet 4.6 on Amazon Bedrock, Google Cloud and Microsoft Foundry. /code-review now reports only bugs and CLAUDE.md violations.

Read ›

GitHub Copilot Evals: repeatable tests for AI agents, now in public preview

GitHub Copilot Evals lets teams write repeatable tests for AI coding skills, custom agents and MCP tools. It checks both the result and the path the agent took, and is in public preview.

Read ›

GitHub Copilot can now drive your browser: browser use enters public preview

GitHub Copilot browser use lets an AI coding agent control Chrome to debug web apps and run browser tests. It is in public preview across the CLI, the Copilot app, VS Code and JetBrains.

Read ›

Claude Code 2.1.327 refuses some symbolic links and lists broken setup in /status

Claude Code 2.1.327 tightens file and package permissions again and adds a /status list of broken agent, skill and hook settings. A short read for teams using AI coding tools on a live product.

Read ›

Start a project

Building with AI? Let's talk.

Tell us what you are building. We'll give you an honest view of the right approach.

Start a projectSee our work