A new arXiv preprint, Plausible Is Not Enough: Trust and Evidence in AI-Authored Pull Request Review, takes up a practical question: do code reviewers hold AI-generated contributions to the same standard as human-written ones? The abstract frames code review as the moment where an AI-authored pull request (AIPR) must earn trust before being merged, and notes that it is unclear whether reviewers apply the same criteria they use for human-authored pull requests (HPRs).
The authors report surveying 239 people, presumably reviewers or developers, but the abstract stops there. No findings, effect sizes, or conclusions are included in the available text. What is clear is the research question: whether the rise of AI-assisted coding changes the tacit rules of review.
The title offers a hint of the likely argument—that a pull request merely looking plausible is not enough, and that reviewers may demand stronger evidence from AI authors. But without the full paper, that remains an inference. Readers should treat the abstract as an introduction to the study's motivation rather than a summary of results.