Does AI-generated detail reduce how much creators own their ideas?
When AI tools produce highly elaborated suggestions, do users feel less personal ownership of those ideas? This matters because felt responsibility may shape how people develop and act on what they create.
In the same 36-participant comparison of ChatGPT against the Oblique Strategies (OS) deck for divergent ideation, the paper's RQ3 and RQ4 findings split quantity and elaboration from ownership. On RQ4, ChatGPT produced ideas with "higher fluency, flexibility, and elaboration" than the OS deck: participants given ChatGPT "generated a greater number of more detailed ideas." On RQ3, the direction reversed on a different axis: ChatGPT users felt "less responsible" for the ideas they produced than OS-deck users did.
The paper does not fully unpack why felt responsibility drops, but ties it to the same property it uses to explain homogenization: "low inferential distance between LLM outputs and apparently finished creative products." If ChatGPT's suggestions already read as complete ideas rather than raw stimuli a person has to finish themselves, users have done less of the work of turning a prompt into an idea — which the paper offers as part of why they claim less ownership of it, even as they accept more of what the model produces and so end up with more and more-elaborated ideas overall.
This is the ownership-side finding from the same experiment that produces Does ChatGPT make different users converge on similar ideas?: the low-inferential-distance property that lets different users converge on similar ChatGPT ideas is also offered as the reason those same users feel less ownership of the ideas they keep. It sits in some tension with Can writers benefit from configuring AI writing partners in advance?, where writers who deliberately plan how they will use an AI partner retain active self-monitoring over the collaboration; here, participants were handed ChatGPT for a one-off ideation prompt with no comparable setup, which may be part of why they report lower ownership of what it returned.
The paper measures self-reported responsibility through a survey administered after each ideation task, not a behavioral or downstream measure of how participants treated the ideas afterward — whether they used, defended, or built on them. The effect is observed only for a single general-purpose chatbot responding to short, researcher-set prompts in a lab session. It does not establish that lower felt responsibility leads to lower-quality follow-through on ideas, only that users who received more and more-elaborated suggestions from ChatGPT claimed less personal ownership of them — a finding that raises, without yet answering, whether people act differently on ideas they don't feel are theirs.
Inquiring lines that read this note 3
This note is a source for these research framings, grouped by the broader line of inquiry each explores. Scan the bold lines of inquiry; follow any specific question forward.
How should control and autonomy be allocated in human-AI collaboration? How should AI ethics integrate consciousness, values, and transparent disclosure? Why do multiple language models independently generate similar outputs?Related concepts in this collection 3
This note in its neighbourhood — explore the map, then jump to a related concept in the list below.
Click a node to walk · click center to open · click Open in graph to see this note in the full knowledge graph
-
Does ChatGPT make different users converge on similar ideas?
When different people use ChatGPT for creative ideation, do they end up with more similar ideas than they would with other tools? This matters because group diversity often drives better collective outcomes.
same 36-participant study's companion finding, on idea diversity rather than ownership
-
Can writers benefit from configuring AI writing partners in advance?
This exploratory study asks whether writers can effectively set up proactive AI assistants by pre-planning their roles and behavior, then use them during actual writing work for idea generation and self-monitoring.
contrasts planned, self-monitored AI collaboration with this study's unplanned, lower-ownership use
-
Do people feel they own AI-generated text they use?
When people use personalized AI to write, do they experience a sense of authorship and ownership? Understanding this matters because it shapes whether disclosure norms around AI use are grounded in how people actually feel.
Evidence for reduced ownership: AI Ghostwriter studies similarly find users don't see themselves as authors of AI-generated text
Related papers in this collection 8
Papers most semantically related to this note, ranked by cosine similarity in the embedding space.
- The AI Ghostwriter Effect: When Users Do Not Perceive Ownership of AI-Generated Text But Self-Declare as Authors
- Homogenization Effects of Large Language Models on Human Creative Ideation
- The LLM Fallacy: Misattribution in AI-Assisted Cognitive Workflows
- Evidence-centered Assessment for Writing with Generative AI
- Has the Creativity of Large-Language Models peaked? —an analysis of inter- and intra-LLM variability —
- The Uneven Impact of Generative AI on Entrepreneurial Performance
- Your Programming Students' Cognition with ChatGPT: Higher Performance, Lower Retention, and Reduced Ownership
- When combinations of humans and AI are useful: A systematic review and meta-analysis
Original note title
ChatGPT users generate more and more detailed ideas but feel less responsible for them than users of a non-AI creativity tool