Tell readers a passage was written by AI and they judge it differently, even when every word is identical.

This explores whether knowing AI was involved changes how we judge a text as we read it, separately from what the text actually says. The corpus's short answer is that the label often matters more than the words. When identical passages were labeled human-written, human judges rated their literary style 13.7 percentage points higher. The surprise is that AI evaluators were 2.5 times more swayed by the same label, a 34.3-point gap. So handing the judging to a model doesn't remove the bias. It makes it stronger Do authorship labels bias how we judge literary quality?.

The penalty isn't the same for every kind of writing. When readers were told a text was AI-authored, they rated it lower on trust, caring and likability. The steepest drops came in personal, interpersonal writing, where readers felt AI couldn't really empathize and treated its use as breaking a social expectation How does revealing AI authorship change reader trust?. This helps explain why readers rate disclosure as more necessary than writers do, especially when the AI text is pasted in directly and couldn't easily be replaced Do readers and writers differ on AI disclosure necessity?. Writers and readers are applying different ideas of what counts as 'written by you.' Writers' own sense of ownership grows with how much they steered the text, not with how personalized the AI was Does user control over AI text shape feelings of ownership?.

This is why the label carries so much weight: without it, people mostly can't tell. A review of 30 studies found that people detect AI-generated text, images and voice at roughly chance levels Can people reliably spot content made by AI?. In practice, then, being told about AI involvement is often the only signal a reader has. Their judgment tracks what they were told more than anything they noticed in the text. Some detectable traces do exist. AI fiction gives itself away through narrative choices: it over-explains themes, prefers tidy single-track plots and avoids moral ambiguity. Readers don't consciously register these traces, but classifiers pick them up Can AI stories be detected without analyzing writing style? Do AI stories explain their themes more than human stories do?.

The less obvious finding is that AI involvement shapes judgments even when readers don't know about it. In a study of nearly 3,000 writers and 11,000 readers, AI assistance shifted how readers saw the writer on all 29 traits measured. Writers came across as more confident, more extreme, more agreeable and more privileged, and no label was needed Does AI writing assistance change how readers perceive the writer?. One explanation for the 'aloof' feel readers report is that human writing quietly appeals for the reader's attention, and AI text doesn't make that appeal, even when a platform shows it to you anyway Does AI writing lack the internal appeal to attention that humans use?.

The bigger picture: we have learned built-in skepticism toward sources like advertising, which we automatically discount as interested speech. We don't yet have a settled stance toward AI-generated text, so it spreads without that filter How do we learn to read AI-generated text critically?. This leaves an odd split. When AI involvement is disclosed, readers react strongly, sometimes too harshly. When it isn't, they get no protective discount at all, even though the text is still changing how they see its author.