The DOI matches its Crossref record; 103 of the 103 references it cites were found in a registry or an index; 19 of the 81 sources its claims cite were checked for support.
Fanfiction in the Age of AI: Community Perspectives on Creativity, Authenticity and Adoption
Open accessAbstract
The integration of Generative AI (GenAI) into creative communities, like fanfiction, is reshaping how stories are created, shared, and valued. This study investigates the perceptions of 157 active fanfiction members, both readers and writers, regarding AI-generated content in fanfiction. Our research explores the impact of GenAI on community dynamics, examining how AI affects the participatory and collaborative nature of these spaces. The findings reveal responses ranging from cautious acceptance of AI’s potential for creative enhancement to concerns about authenticity, ethical issues, and the erosion of human-centered values. Participants emphasized the importance of transparency and expressed worries about losing social connections. Our study highlights the need for thoughtful AI integration in creative platforms using design interventions that enable ethical practices, promote transparency, increase engagement and connection, and preserve the community’s core values.
Overview
CoChat Summary
TL;DR
This survey study examines how 157 active fanfiction readers and writers perceive generative AI in fanfiction, finding a spectrum from cautious acceptance to strong concern about authenticity, ethics, transparency, and community erosion. Participants largely value fanfiction as a human creative and social space, while some writers already use AI tools for limited tasks.
Key findings
- Most participants saw fanfiction as a space for human creativity (92%, 145/157), and many doubted AI could replicate emotional nuance and depth (84.7%, 133/157) or maintain authenticity while offering innovative storytelling (77.5%, 121/156).
- Transparency was central: 86% (135/157) said authors should disclose AI involvement; 66% (103/156) said knowing a story was AI-generated would decrease their interest, and 72.2% (112/155) reported negative feelings if they discovered after reading that a fic was AI-written, with 58% (90/155) feeling deceived.
- Writers were divided: 74.4% (67/90) reported using at least one AI tool, 33% (30/90) never used any, and 10% (9/90) supported AI fully generating fanfiction stories; more experienced writers (>10 years) were significantly more resistant to higher AI involvement than early-career writers (1–5 years).
- Concerns included AI posing a danger to the social aspects of the community (76.4%, 120/156), inundating platforms with AI-created stories (83.4%, 131/157), stifling human creativity (79.6%, 125/157), loss of personal touch/authenticity (78.2%, 122/156), low-quality content (73.7%, 115/156), and ethics (68.6%, 107/156).
Methods
- Structured online questionnaire distributed via email lists, Dreamwidth communities, private Discord servers, and social media; conducted May–July 2024, with questions for readers, writers, or all participants.
- Initial 489 respondents were cleaned to 157 after removing 221 bot responses from one location and 111 additional suspected bots; final sample included 90 writers and 67 exclusive readers, with one writer identifying as a non-reader.
- Quantitative analysis used Python, SPSS, and R, including Mann-Whitney U and chi-square tests; open-ended responses (Q29, Q36, Q46) were content-analyzed by two independent coders with agreement rates of 86.6%, 99.3%, and 86.3%, respectively.
Limitations
- Stated: the sample is skewed toward North American and English-speaking participants, limiting generalizability to global fanfiction communities, especially non-English contexts; the authors call for more regionally and linguistically diverse participants.
- Stated: the study captures a transition phase rather than a longitudinal view; the authors say future work should use longitudinal methods and in-depth interviews to trace evolving norms, attitudes, and covert/overt AI use.
- Evident from the text: the findings rely on self-reported survey responses and are cross-sectional, so they cannot establish causal effects of AI integration on community dynamics, creativity, or skill development.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures










Verification
Claims the paper makes
16 Valid · 3 Not valid · 46 Not checked · 165 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“online spaces are seeing an increased presence of overt and covert intelligent agents that are altering the communities' culture and dynamics”
§ Community Perspectives on Creativity, Authenticity and Adoption > 1. Introduction
What the sources say
- Supportedcites Tsvetkova et al., 2024 source
The Abstract describes bots and algorithms proliferating across communication channels and social interactions. Box 2 describes covert Twitter bots changing discourse, while Box 4 describes evident Reddit bots influencing users’ language and sentiment. Together, these passages support the paraphrase about increasing presence and effects on online communities. (LLM, conf=0.89)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Recent investigations into creative communities have revealed widespread apprehension regarding the unauthorized use of artistic works in AI training (Kyi et al., 2025;Lovato et al., 2024).”
§ Community Perspectives on Creativity, Authenticity and Adoption > 6.2.3. Preserving community values
What the sources say
- Supportedcites Kyi et al., 2025 source
The Abstract (p. 1) states that since generative AI's emergence, creative workers have spoken up about career-based harms, with 'a common theme in these accounts of harm is that generative AI models are trained on workers' creative output without their consent and without giving credit or compensation'. Our copy's empirical section corroborates this at the level of the sample: 4.2.3 (p. 8) is headed 'Generative AI models being trained on copyrighted content was a common concern', and 5.2 (p. 10) reports creatives 'worries about losing control over their work, such as it being used for training without their knowledge or consent'; 4.1.1 (p. 5) records that almost all participants wanted to be asked for consent. The apprehension is specifically about unauthorised (non-consented) use, which matches the claim; the only narrowing is that this study's sample is 20 creative workers in visual art/design, writing and programming rather than creative communities at large, though the paper itself frames the concern as a common theme across creative workers worldwide (2.3, p. 3). (LLM, conf=0.85)
- Supportedcites Lovato et al., 2024 source
Lovato et al. (2024) is a survey of 459 artists (Results, p. null) whose findings the authors summarise as 'a strong sentiment among artists in favor of disclosing what art is being used in AI training ... and expressing concerns regarding AI's impact on the art industry's workforce and who profits from their art', with 80.17% requiring detailed disclosure and 61.87% agreeing AI models threaten art workers (Introduction/Results, p. null). The Abstract frames the grievance precisely as non-consensual use: 'Without explicit consent from artists, Generative AI creators scrape artists' digital work to train Generative AI models', and the Economic Context section states that harm 'can arise from the use of content without explicit consent' and that this concern 'is shared among artists, creative industries, and government agencies alike'. The one qualification worth noting is that the apprehension documented is about undisclosed/for-profit exploitation rather than training use per se ('artists are not wholesale against art being used in AI art models training data'), which still matches 'apprehension regarding the unauthorized use'. Population is US-recruited artists, narrower than 'creative communities' generally, but the claim's breadth is also carried by the co-cited Kyi et al. (LLM, conf=0.85)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“As Jo and Gebru (2020) and Lovato et al. (2024) note, questions of agency, visibility, and ownership are not evenly distributed within these sociotechnical systems.”
§ Community Perspectives on Creativity, Authenticity and Adoption > 6.4. Theoretical implications
What the sources say
- Supportedcites Jo and Gebru (2020) source
Section 5.2 describes how researcher-defined labels constrain participants’ input, while Section 5.3 explicitly discusses power inequities and unequal data ownership. The Consent and Transparency sections address underrepresentation and what data-collection decisions are made visible. Together, these support the broad paraphrase in the context of ML data collection; page numbers are unavailable in our copy. (LLM, conf=0.88)
- Supportedcites Lovato et al. (2024) source
The “Problems in the Pipeline of Generative AI Tools” section describes disparities in who creates, profits from, and uses AI tools, alongside companies’ secrecy about how the tools work. The Results and Discussion sections show that ownership of AI-generated artwork is contested. Together, these passages support the claim about uneven agency, visibility, and ownership in these systems. (LLM, conf=0.88)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Ilonka Gero et al. (2025) observe that fanfiction authors often experience this form of AI-driven repurposing as misrecognition or disempowerment, particularly when their stylistic voice is mimicked or reassembled without consent or attribution.”
§ Community Perspectives on Creativity, Authenticity and Adoption > 6.4. Theoretical implications
What the sources say
- Partly supportedcites Ilonka Gero et al. (2025) sourceExperimental — accuracy not yet measured
Section 5.1.2 (page unavailable) reports that a fan-fiction writer found putting deeply personal work into a language model disturbing, and Section 5.2.1 reports writers’ lack of control over such use. The paper also discusses mimicked style and uncredited words as concerns among participants, but it does not establish that fanfiction authors often experience AI mimicry of their own voices as misrecognition or disempowerment. (LLM, conf=0.86)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This can be seen through the lenses of forming an identity within the writers' domain, as writers with an established writing identity need to engage more in the sense-making of the change and its impact compared with novices (Schmitt et al., 2024).”
§ Community Perspectives on Creativity, Authenticity and Adoption > 6.3.2. Adoption of GenAI tools
What the sources say
- Partly supportedcites Schmitt et al., 2024 sourceExperimental — accuracy not yet measured
The Introduction says software engineers’ sense-making about generative AI differs with domain expertise, and §5.1.1 describes established occupational identity as potentially threatened by new technology. The paper studies software engineers, not writers, and does not establish that experienced workers need to engage in more sense-making than novices. (LLM, conf=0.84)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
112 more claims
Create a free account to see all 230 claims and every source verdict.
- Probably not a substantial claim. 81
- It may state two separate claims as one. 62
- The quote only partly states this claim. 3
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“When asked about use rather than familiarity, 74.4% (67/90) reported using at least one of the AI tools, and 33% (30/90) reported never using any AI tools.”
§ Community Perspectives on Creativity, Authenticity and Adoption > 4.2.1. Fandoms and platforms
Reported 90 · recomputed 97
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“14% have secondary education (24/157)”
§ Community Perspectives on Creativity, Authenticity and Adoption > 4.1. Participants demographics
Reported 14 · recomputed 15.286624
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“40.9% (36/90) of the writers did not find it played any role”
§ Community Perspectives on Creativity, Authenticity and Adoption > 5.1.2. Writers' views on AI use in their creative process
Reported 40.9 · recomputed 40
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“Education: (Q4) 55.4% (87/157) of the participants have an undergraduate degree, 25.48% (40/157) have a post-graduate degree, 14% have secondary education (24/157), 6 have vocational or technical training (e.g., apprenticeships, trade schools), 2 are undergrad students, and one preferred not to say.”
§ Community Perspectives on Creativity, Authenticity and Adoption > 4.1. Participants demographics
Reported 157 · recomputed 160
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“64.74%, 101/159”
§ Community Perspectives on Creativity, Authenticity and Adoption > 4.2.3. Engagement motivation
Reported 64.74 · recomputed 63.522013
1 more finding
Create a free account to see every finding of this check.
Citations
Cites
Valente, T. W. (2003) · Genetics in Medicine · doi:10.1097/01.gim.0000061743.67794.c4 · Cited by 13,723
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988
Convergence Culture: Where Old and New Media Collide
Jenkins, H., & Plasencia, A. (2017) · Is the Universe a Hologram? · doi:10.7551/mitpress/9780262036016.003.0012 · Cited by 6,944
Convergence Culture: Where Old and New Media Collide
Carbonell, C. (2007) · The Journal of Popular Culture · doi:10.1111/j.1540-5931.2007.00435.x · Cited by 5,530 · cited 3×
“Still, despite these technical proposals, many creators report feeling side-lined from the decisions that shape how their contributions are used, raising questions about representation, legitimacy, and the governance of creative labor in AI development [56, 62].”
Ray, P. P. (2023) · Internet of Things and Cyber-Physical Systems · doi:10.1016/j.iotcps.2023.04.003 · Cited by 2,463 · cited 1×
“Recent discussions in AI ethics increasingly emphasize that creative output should not be treated as neutral raw material, but rather as the outcome of human labor embedded in cultural, social, and emotional contexts [52].”
Data colonialism: rethinking big data’s relation to the contemporary subject
Couldry, N., & Mejias, U. A. (2019) · Television & New Media · doi:10.1177/1527476418796632 · Cited by 1,617 · cited 2×
“This includes tracing shifts in attitudes toward authorship, originality, and attribution, as well as concerns around consent, transparency, and the legitimacy of AI-assisted contributions.”
(2016) · Open Research Online (The Open University) · Cited by 1,410
Generative artificial intelligence enhances creativity but reduces the diversity of novel content
Doshi, A., & Hauser, O. (2023) · doi:10.2139/ssrn.4535536 · Cited by 1,350 · cited 6×
“To address these challenges, researchers emphasize the need for user-centered design of more effective AI tools for creative writing [59, 104].”
InfluentialThe Costs of Connection: How Data Is Colonizing Human Life and Appropriating It for Capitalism
Couldry, N., & Mejias, U. A. (2020) · doi:10.1515/9781503609754 · Cited by 1,246
Algorithms as culture: Some tactics for the ethnography of algorithmic systems
Seaver, N. (2017) · Big Data & Society · doi:10.1177/2053951717738104 · Cited by 1,010 · cited 1×
“Rather than viewing generative AI as a neutral technological tool, it is more appropriate to conceptualize it as a cultural system embedded with human assumptions, institutional agendas, and structural asymmetries [12, 79].”
Research works printed but not linked: 7
- Black, R. W. (2008). Adolescents and online fan fiction (Vol. 23). Peter Lang.
- De Cremer, D., Bianzino, N. M., & Falk, B. (2023). How generative AI could disrupt creative work. Harvard Business Review, 101(2), 102–111. https://hbr.org/2023/04/how-generative-ai-could-disrupt-creative-work Accessed: 2023-08-22.
- 10.55554/27859649.1017 · doi:10.55554/27859649.1017
- Khan, S. (2024). Brave new words. Penguin Random House.
- Prakash N. (2024). The empire types back: Colonialism resistance and storytelling online [Ph.D. Dissertation]. University of Oxford.
- Rouse L. & Stanfill M. (2023). Over*Flow: Fan demographics on archive of our own. Flow Journal. https://www.flowjournal.org/2023/02/fan-demographics-on-ao3/
- Shope, M. (2023). Best practices for disclosure and citation when using artificial intelligence tools. Georgetown Law Journal Online, 112, 1–22. https://ssrn.com/abstract=4338115.
Cited by
Mind Your Manners: The Dynamics of Politeness in Human-AI vs. Human-Human Interactions
Lazebnik, T., Zalmanson, L., & Mokryn, O. (2025) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3757631 · Cited by 11
“The socio-demographic questions were designed to capture a wide range of participant characteristics that could influence their interaction with conversational AI tools from previous studies [2, 33, 44, 45, 53, 66].”
Self-citationResponsible and sustainable use of AIGC image tools in art education: an SEM–ANN–fsQCA study
Xia, B., Luo, Y., & Su, Y. (2026) · Education + Training · doi:10.1108/et-05-2026-0609
Cui, J. (2026) · International Journal of Cultural Studies · doi:10.1177/13678779261448882
Metadata
- Authors
- Alfassi, R., Cooper, A., Mitchell, Z., Calabro, M., Shaer, O., & Mokryn, O.
- Affiliations
- University of Haifa (IL)
- Wellesley College (US)
- Corresponding author
- Mokryn, Osnat
- Venue
- International Journal of Human–Computer Interaction
- Publisher
- Informa UK Limited
- Published
- 2025-08-12
- Volume
- 42
- Issue
- 5
- Pages
- 3062-3094
- ISSN
- 1044-7318 (ISSN-L)
- 1532-7590
- OpenAlex
- W4413197805
- Semantic Scholar
- c71471af6806524e978d3b2ca443be7e3bc22d2c
- Cited by
- 4
- References
- 86
- Citations by year
- 2026: 3 · 2025: 1
- FWCI
- 4.12
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- International Journal of Human-Computer Interaction (publishedVersion, cc-by-nc-nd)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Law › Literary and Philosophical Studies
- Social Sciences › Communication › Literature, Culture, and Aesthetics
- Keywords
- fan fiction, generative AI, AI-generated content, creative communities, authenticity, community dynamics, collaborative creativity, transparency, ethical considerations, social connectedness, design interventions, creative platform
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-12, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-12 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Fanfiction in the Age of AI: Community Perspectives on Creativity, Authenticity and Adoption. https://app.cochat.ai/papers/10.1080/10447318.2025.2531272
- The paper
- Alfassi, R., Cooper, A., Mitchell, Z., Calabro, M., Shaer, O., & Mokryn, O. (2026). Fanfiction in the Age of AI: Community Perspectives on Creativity, Authenticity and Adoption. International Journal of Human–Computer Interaction. https://doi.org/10.1080/10447318.2025.2531272
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.