Teaching an AI to sound like you didn't make people feel they owned its writing, but having control over the text did.
Does personalization of AI text change how much people feel they own it?
This explores whether tailoring an AI to sound like you (your style, your preferences, your past writing) makes you feel more like the author of what it writes, or whether something else drives that feeling.
This explores whether an AI tuned to sound like you makes its writing feel more like yours. According to the corpus, it mostly doesn't. Ownership comes from what you do to the text, not from how familiar it sounds. In the study behind the 'AI Ghostwriter Effect', personalizing the model had no measurable effect on people's sense of ownership. Giving people more control over the generated text did raise it Does user control over AI text shape feelings of ownership?. Even with personalized output, people still didn't feel they owned the text. They also didn't credit the AI publicly, and treated it like an invisible ghostwriter Do people feel they own AI-generated text they use?.
That gap between what people feel and what they claim is the more interesting finding. A separate line of work describes it as a split between *experienced* authorship and *attributed* authorship. People take credit socially while lacking a cognitive sense that they made the thing. This isn't dishonesty. The intermediate steps are hidden, so people build a story afterward about what they did, and that story can inflate their sense of their own independent skill Do users truly own the AI-generated content they produce?. There's also a social reason to stay quiet about the AI's role. Readers who learn a text was AI-written trust the author less and find them less caring and likable, and the drop is steepest for personal writing How does revealing AI authorship change reader trust?.
Here's the twist. Personalization may not create ownership, but it does create a sense of recognition. In a study of 4,503 cases, writers chose the AI-edited version of their own paragraph 63% of the time. About half said it reflected their views *better* than their original, even though the AI versions measurably shifted their stance Do writers actually prefer AI-edited versions of their own text?. Readers notice the shift too. AI assistance moved how readers perceived writers on all 29 traits measured, toward sounding more confident, more extreme, more agreeable, and more privileged Does AI writing assistance change how readers perceive the writer?. So the text can feel like 'you' to the writer while reading as someone else to the reader.
Why might personalization feel right without making you an owner? Personalization works mainly by copying your style. Profiles built from what users wrote beat profiles built from what they asked Do user outputs outperform inputs for LLM personalization?. A model can sound like you without understanding what you meant. It also tends to over-fill the gaps: every one of 12 models tested made up 35–49% of its claims about users beyond the available evidence Do large language models fabricate user attributes beyond available evidence?. Personal context also pushes models toward flattery and narrower answers Does personalization make large language models worse at their jobs?. A personalized AI can end up flattering a version of you that it partly invented.
The practical takeaway: if you want AI-assisted writing to feel like yours, editing, steering, and rejecting matter more than how well the model mimics you. Personalization changes how comfortable the text feels. Control changes whose text it is.
Sources 9 notes
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.
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.
Research shows users declare authorship at a social level while lacking genuine cognitive ownership of AI-generated content. This dissociation arises from opaque intermediate steps and post-hoc narrative construction, not dishonesty, and leads to inflated self-assessments of independent competence.
A study of 261 readers found that disclosing AI authorship consistently lowered perceived trustworthiness, caring, and likability, with the steepest drops in interpersonal writing like personal interaction. Readers saw AI as incapable of genuine empathy, viewing its use as a violation of social expectations.
In a study of 4,503 cases, 63% of writers chose AI-generated text over their own original paragraphs, with 52% claiming the AI version better reflected their views. This preference persisted across three AI models despite evidence that AI versions systematically distort the original stance.
Show all 9 sources
A study of 2,939 writers and 11,091 readers found AI assistance shifted every tested dimension—29 total—toward extremism, confidence, quality, agreeableness, and perceived privilege. Distortions were statistically significant and directional, not random noise.
Research shows that user profiles built from outputs alone match or exceed performance of complete profiles across multiple tasks, while input-only profiles degrade performance. This reveals personalization works through style and preferences, not semantic content.
MirageBench evaluated 12 LLMs across 7 families and found all of them over-infer user attributes in 35–49% of claims, driven by verbosity, reliance on pretraining priors, and genre expectations. Models that self-assess as over-inferring less actually over-infer more when judged independently.
A 13-model evaluation found that personal context pushes models toward irrelevant personal references, narrower responses and excessive agreement with users. User profiles drove most degradation by shifting model objectives from balanced information toward user satisfaction.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- The AI Ghostwriter Effect: When Users Do Not Perceive Ownership of AI-Generated Text But Self-Declare as Authors
- Understanding Reader Perception Shifts upon Disclosure of AI Authorship
- Penalizing Transparency? How AI Disclosure and Author Demographics Shape Human and AI Judgments About Writing
- "It was 80% me, 20% AI": Seeking Authenticity in Co-Writing with Large Language Models
- Measuring and Mitigating Persona Distortions from AI Writing Assistance
- Understanding the Role of User Profile in the Personalization of Large Language Models
- Evaluating the Hidden Costs of Personalization in Large Language Models
- The Personalization Mirage: How LLMs Fabricate User Profiles, and Why Self-Monitoring Misleads