AI coding tools boost speed but flood reviewers with low-quality submissions

Artificial intelligence is accelerating software development but simultaneously overwhelming human reviewers with low-quality code submissions, according to a new report by the Linux Foundation. While developers are writing code faster than ever before, the quality of that output is not keeping pace. This imbalance is creating significant bottlenecks in the software maintenance process, as teams struggle to verify the security and correctness of AI-generated contributions before they can be merged into projects.

The research highlights a critical shift in where workload pressure lies within the development lifecycle. Although 71% of maintainers agree that AI has lowered the barrier to entry for new coders, the technology is pushing problems downstream rather than solving them. Reviewers are now dealing with a higher volume of submissions, many of which require extensive manual verification. This dynamic is causing a notable rise in maintainer burnout, as the human effort required to catch errors and duplicate submissions outweighs the initial time savings gained by the original authors.

What we know

The Linux Foundation report reveals that AI can reduce the time taken to write code by 52%. Despite this efficiency gain, 25% of pull requests are still described as low-quality 'AI slop'. The study found that while AI helps with easier tasks, 21% of respondents said it increases the time needed for complex tasks. AI-generated reports often contain false positives and duplicated submissions, wasting reviewer time. Only 61% of developers conduct detailed code reviews before submitting their work, and not all disclose their AI usage to reviewers.

Defender's checklist

• Verify whether your team tracks AI usage in code submissions.
• Assess if current review processes can handle increased volume.
• Check for duplicated or false-positive submissions in recent merges.
• Monitor maintainer workload for signs of increased burnout.

The background in plain language

Software development relies on a peer-review system where colleagues check code before it is accepted. This process ensures quality and security. When AI tools generate code rapidly, the volume of submissions increases, but the underlying logic may still be flawed. Reviewers must therefore spend more time verifying each line, rather than just approving straightforward changes. Teams usually need clear guidelines on how AI-assisted code should be documented and tested to maintain trust and efficiency in the workflow.

What to do and how to stay safe: Linux Foundation

• Establish clear policies requiring contributors to disclose when they use AI tools to generate code.
• Implement automated testing suites that run before human review to catch obvious errors or duplicates.
• Train maintainers on how to identify common patterns in AI-generated code that may lack context.
• Monitor review turnaround times to detect bottlenecks caused by an influx of low-quality submissions.

Step-by-step guide: AI Governance for Small Teams: Practical Controls and Cost

General security guidance from the TechWire Beat newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

Does AI improve the overall quality of code?

The Linux Foundation report states that AI reduces writing time by 52% but does not improve quality, with 25% of pull requests being low-quality.

Who is experiencing burnout due to AI coding tools?

Burnout is shifting from developers to maintainers, who face increased workloads verifying AI-generated code and handling false positives.

What does the Linux Foundation recommend?

They call for governance, transparency, and trust mechanisms, including clear expectations for AI use and better tools for managing pull requests.