If you sign AI-drafted text as your own, do you feel it's yours, or just a ghostwriter's work?
Do writers claim authorship without feeling they wrote the words?
This explores the gap between what writers publicly claim and what they privately feel when AI drafts the words: do people put their name on AI-generated text while knowing, inside, that they didn't really write it?
This explores the gap between public authorship and private ownership: whether people sign AI-drafted text as their own while not feeling it's theirs. The short answer from the collection is yes, and researchers have named it the AI Ghostwriter Effect. In two studies, people said they did not feel they owned text an AI wrote for them, yet they also declined to credit the AI when presenting it publicly. They treated the model like a hired ghostwriter whose name never appears Do people feel they own AI-generated text they use?. So felt authorship and declared authorship split apart.
What closes that gap? Not making the AI sound more like you. Personalizing the model to a user's own style had no effect on how much they felt they owned the output. What mattered was control: the more a person shaped and steered the generated text, the more it felt like theirs Does user control over AI text shape feelings of ownership?. Interviews with professional writers point the same way. They place authenticity in the act of making the text (where the ideas came from, who they are, what it was like to build it) rather than in the finished words Where do writers locate authenticity in AI co-writing?. Ownership comes from doing the work, not from the output resembling you.
Here's a twist: telling people they own the result can increase how much they hand off. Writers told they owned the final product leaned more heavily on AI suggestions. Writers told they were composing their own work spent more effort revising it themselves Does ownership framing change how much writers rely on AI?. Being told 'this is yours' doesn't make people write more of it. It can license them to write less.
Why stay silent about the AI? Readers mostly can't tell AI-assisted writing from solo writing Do readers value writing authenticity they cannot detect?, and heavy rewriting may erase the traces further, though that claim hasn't been directly tested Do rewrites that hide authorship also fool AI detectors?. Labels, though, do carry weight. The same passage gets rated noticeably higher when labeled human-written, and AI judges show this bias even more strongly than people do Do authorship labels bias how we judge literary quality?. Writers who never used AI can still be accused of using it, so suspicion lands on them anyway Do unfounded AI accusations harm human writers instead?. Given all that, keeping quiet looks like a rational defense.
The surprise is that the fear behind the silence may be misplaced. In one study of scientific abstracts, readers' judgments followed what they believed about AI involvement more than the actual authorship. And openly disclosing authorship raised trust and quality ratings across every type of abstract, reversing the penalty earlier work had found Do reader judgments reflect actual authorship or just their beliefs?. Writers may be hiding the ghostwriter to avoid a penalty that being open could actually prevent.
Sources 9 notes
Two studies (n=30, n=96) found users do not feel they own AI-generated text, yet they refrain from publicly crediting the AI—treating it like an invisible ghostwriter. This gap between felt and declared authorship held even when AI text was personalized.
Study 1 found that greater user control over generated text raised sense of ownership, while personalizing the AI model had no impact on the AI Ghostwriter Effect.
Professional writers co-writing with AI emphasize internal experience and the act of construction as central to authenticity, beyond the resulting text. Interviews with 19 writers revealed they define authenticity through source, identity, and the lived experience of making.
Writers told they own the final product relied significantly more on AI suggestions, while those framed as composing their own work focused on self-revision. This ownership effect shaped the writing process independent of AI quality.
Hwang et al. found that readers could not distinguish AI-assisted from solo-written work and showed positive attitudes toward AI use. However, the study did not test whether readers would value process authenticity if disclosure occurred or if they could perceive it.
Show all 9 sources
The paper asserts that rewritten messages evade AI-text detectors but provides no detector experiments, only attribution results showing stylistic convergence. The double erasure claim needs direct empirical testing.
Human judges rated identical passages 13.7 percentage points higher when labeled human-authored; AI models showed a 2.5-fold stronger bias at 34.3 points. The effect persists across AI architectures, suggesting evaluators respond to provenance cues rather than text quality alone.
Accused comments lack features that distinguish AI text from human writing, suggesting accusations function as gatekeeping rather than detection. This inverts the AI-as-perpetrator framing, placing harm at the receiving side through reader skepticism.
Readers' evaluations of abstracts were shaped by their beliefs about LLM involvement rather than actual authorship. Crucially, disclosing authorship raised trust and quality ratings across all abstract types, reversing the credibility penalty shown in prior work.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- Understanding Reader Perception Shifts upon Disclosure of AI Authorship
- Penalizing Transparency? How AI Disclosure and Author Demographics Shape Human and AI Judgments About Writing
- Measuring and Mitigating Persona Distortions from AI Writing Assistance
- The AI Ghostwriter Effect: When Users Do Not Perceive Ownership of AI-Generated Text But Self-Declare as Authors
- What Influences Readers' and Writers' Perceived Necessity of AI Disclosure?
- "It was 80% me, 20% AI": Seeking Authenticity in Co-Writing with Large Language Models
- The human-authorship halo: attribution bias in literary style evaluation by humans and AI
- LLM or Human? Perceptions of Trust and Information Quality in Research Summaries