The claim

“Code written with AI assistants contains more security vulnerabilities than code written without them.”

Began as a single source, before casesByPsamiKoinovera teamPublished 5 Oct 2026
Supports

Security Weaknesses of Copilot-Generated Code in GitHub Projects: An Empirical Study

The source shows that an analysis of Copilot-generated snippets found in public GitHub projects identified security weaknesses in a substantial share, across dozens of weakness types.[1]

[1] Security Weaknesses of Copilot-Generated Code in GitHub Projects: An Empirical StudyAcademic Paper

arXiv.org · 6 Feb 2025

Passage from the sourceOur analysis identified 733 snippets, revealing a high likelihood of security weaknesses, with 29.5% of Python and 24.2% of JavaScript snippets affected.

Rate or review this source

How readers rated the reasoning

These rate the argument. Each source is rated separately, on the source. Neither is a verdict on the claim.

No one has rated this case yet.

Review notes

Short, specific corrections to this case’s reasoning. A full opposing argument belongs in a case of its own.

Sign in to add note

Loading review notes…

Gathering the latest community-submitted evidence.