--- title: "AI polish makes it harder to spot problems. But there’s a quick fix." url: https://daily.dev/posts/ai-polish-makes-it-harder-to-spot-problems-but-there-s-a-quick-fix--pndonfylc source_url: https://www.atlassian.com/blog/ai-at-work/ai-polish-makes-it-harder-to-spot-problems-but-theres-a-quick-fix type: article source: "Atlassian" published: 2026-08-25T22:35:38.377Z updated: 2026-08-25T22:41:08.427Z tags: ["genai", "code-review"] reading_time: 5 upvotes: 0 comments: 0 language: 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. # AI polish makes it harder to spot problems. But there’s a quick fix. **[Atlassian](https://daily.dev/sources/atlassian)** · 5 min read · 0 upvotes · 0 comments ## Summary Atlassian's Teamwork Lab ran a controlled study with 903 knowledge workers to test whether AI-polished writing makes it harder to catch substantive flaws. Reviewers of AI-polished drafts were 22% less likely to spot a first flaw and 15% less likely to spot a second, spent 62% more time reviewing, and were 18% less willing to give critique. However, adding a simple 'Status: Early Draft' label almost entirely reversed these effects while keeping perceived effort ratings high. Older employees and managers proved largely immune to the polish effect, while Gen Z and entry-level workers showed the steepest drops in flaw detection. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Does polishing a draft with AI make reviewers less likely to catch mistakes in it? Yes, reviewers evaluating AI-polished drafts were 22% less likely to catch a first flaw and 15% less likely to catch a second flaw compared to reviewers of unpolished drafts, according to a controlled study of 903 knowledge workers. They also spent 62% more time reading and were 18% less willing to give critique, since polish signals the work is already finished. _Anyone relying on AI-polished writeups can track workflow research like this on daily.dev before it burns them in review._ ### How do you get honest feedback on a document that was polished with AI? Add a simple label like 'Status: Early Draft' along with an explicit request for feedback. In a study of 903 reviewers, this almost entirely negated the AI-polish effect, dropping perceived finality from 37% to 21% and restoring willingness to critique to baseline, while high-effort ratings actually rose from 64% to 69%. _Teams refining AI-assisted writing workflows can follow findings like this on daily.dev._ ### Are some employees better than others at catching flaws in AI-polished writing? Yes, Gen X employees (46+) and managers acted as a natural firewall, spotting flaws in proposals regardless of whether AI polish was applied. Gen Z and entry-level employees, by contrast, showed steep drops in flaw detection when reviewing AI-polished drafts versus unpolished ones. _Managers weighing how AI writing tools affect review quality can follow research like this on daily.dev._ ## Similar posts on daily.dev - [AI workslop: The golden touch that’s killing productivity](https://daily.dev/posts/ai-workslop-the-golden-touch-that-s-killing-productivity-dwj2xvpob) · SD Times · 1 upvotes · 0 comments --- Tags: [#genai](https://daily.dev/tags/genai), [#code-review](https://daily.dev/tags/code-review) [View this post on daily.dev](https://daily.dev/posts/ai-polish-makes-it-harder-to-spot-problems-but-there-s-a-quick-fix--pndonfylc)