Shuichiro Ogawa
日本語

Notes · updated 2026-08-02

Which Angle for a Generative-Art Review Paper Attracts Citations? An Empirical Analysis Using Bibliographic Data

Scope and Method

This note empirically identifies, with public bibliographic data, the “content and angle most likely to attract citations” when writing a review paper on generative-art research. Building on the previous literature map generative-art-literature (a 94-item corpus), it collects and analyzes the following four questions. An adversarial review of this note’s ranking logic is in generative-art-adversarial-review.

  • Question 1 (R series, 43 items): measured citation counts and citation velocity of existing review/survey papers.
  • Question 2 (G series, 14 data points): growth comparison of publication counts and citations across candidate themes (2020–2026).
  • Question 3 (S series, 9 items + M series, 7 items): on-the-ground confirmation of structural elements common to highly cited surveys, and scientometrics meta-research on citation determinants.
  • Question 4 (X series, 6 items): gap areas lacking comprehensive reviews, and the evidence for their citation demand.

Citation counts were measured with the OpenAlex API (peer-reviewed journal DOIs) and the Semantic Scholar API (arXiv preprints), and every figure carries its retrieval date (2026-08-02). Google Scholar figures are not used because they cannot be verified via API. Citation velocity (cited_by_count divided by years since publication) serves as the primary metric for comparison. The corpus is source/review/generative-art-review-citation-strategy/papers.md.

The method’s limits come first. Theme-level publication counts depend on the co-occurrence of search terms, and counts vary greatly across synonyms (834 items for “human-AI co-creation” versus 90 for “co-creativity”). Because 2026 is a partial count as of August, growth judgments are anchored on 2024→2025. The judgment that “no review exists” is a qualified claim: none could be confirmed within the range of the recorded search terms.

Question 1: What Earns Citations for Existing Reviews

The measured citations (R series, 43 items) reveal a four-tier hierarchy of citation velocity.

First, the fastest tier is large technical surveys. The diffusion-model survey (Yang et al. 2023) reached 1,441 citations in three years (288.2 per year), and the AIGC survey (Wu et al. 2023) reaches 101.8 per year. However, the readership of this tier is the CS/ML community, which does not overlap with the readership of a generative-art research review. Competition is also already saturated (recorded as a comparison baseline).

Second, empirical studies from 2023–2025 rank next in citation velocity after technical surveys. Zhou & Lee 2024, which estimated the effect of text-to-image adoption over roughly four million works, reached 383 citations in two years (191.5 per year); Oppenlaender et al. 2024, which verified prompting as a skill, reaches 85.5 per year; and Bellaiche et al. 2023, the preference experiment on AI labels, reaches 51.5 per year. That primary studies, not reviews, are cited at this speed is because common empirical results to reference are in short supply, which serves as circumstantial evidence of review demand in this area.

Third, experimental psychology of AI-art reception published in 2018–2019 continues to be cited at around 25 per year (Chamberlain 2018, Hong & Curran 2019, and Mazzone & Elgammal 2019 total 468 cumulative citations).

Fourth, theory and evaluation frameworks of computational creativity maintain a stable base of 14–30 per year. Boden 1998 (798), Boden 2009 (518), Colton & Wiggins 2012 (344), Ritchie 2007 (288), Wiggins 2006 (285), and SPECS (Jordanous 2012, 205) exceed 3,850 cumulative citations. They do not explode, but they have not withered either. The existence of this “stably and continuously cited body of theory” grounds the number-one angle in the ranking below.

By contrast, reviews confined to generative art (the traditional term) show no citation growth. Galanter 2016 has 60 citations in ten years (6.0 per year), and McCormack et al. 2014’s “ten questions” stays at 6.5 per year. The exception is the conceptual analysis of McCormack et al. 2019 (four concepts of authorship, 175 citations, 21.9 per year), whose citations accelerated once connected to the authorship controversy around generative AI. Whether one writes within the older frame or connects to the current controversy makes a threefold difference in citation velocity even for the same author group.

Question 2: Which Themes Are Surging in 2024–2026

The growth comparison of publication counts (2024→2025 growth rates) yields the following order.

RankTheme24→25 growth2025 publications22→25 multiplier
1human-AI co-creation+607% (42→297)29729.7x
2Generative AI x copyright (IP-narrowed)+160% (50→130)1308.7x (23→25)
3LLM creativity evaluation+115% (131→282)28294x
4AI-art reception and perception+83% (23→42)4210.5x
5generative art (the term itself)+57% (95→149)1493.5x
6text-to-image evaluation+47% (1,326→1,943)1,9436.5x
Ref.diffusion models overall+40%17,8385.9x

The reading requires care. The +607% for human-AI co-creation is the fastest, but the term’s surge also means fast review competition (the HCI systematic review by Hu et al. 2025 has 26 citations one year after publication, and the scoping review by Tsao et al. 2025 already exists). Conversely, text-to-image evaluation is large in scale but low in growth rate, entering a mature phase. Combining scale, growth, and competition, the target is an area that is “fast-growing, mid-sized in scale, and not yet reviewed,” which the gap analysis of Question 4 identifies.

Review demand as a whole is also expanding. Review-type publications related to generative art grew 6.6-fold, from 136 in 2022 to 897 in 2025.

Question 3: The Structure of Highly Cited Surveys and General Findings on Citation Determinants

On-the-ground confirmation of the structural elements of nine highly cited surveys (within the range where full text could be confirmed) yields the following frequencies.

  • Taxonomy (classification scheme): 7/7. Present in all.
  • Independent open problems / future directions section: 7/7. Present in all.
  • Organized definitions: present in all four that could be confirmed.
  • Comparison tables (quantitative comparison of benchmarks/methods): 3/4.
  • Genealogy or timeline figures: 3/5.
  • Accompanying GitHub repository: 2/6. Only the top two by citations (Yang 2023 at 1,441 and the Zhao et al. LLM survey) have one.
  • Explicit systematic methodology such as PRISMA: 0/6. CS-style surveys do not use PRISMA (a clear difference from medicine).

Taxonomy and open problems are close to necessary conditions, and a GitHub repository (an awesome-list-style bibliography) is the common trait of the topmost tier. A branching is also observable: taxonomy over PRISMA when targeting CS readers, and PRISMA-ScR (adopted by Tsao 2025 and Hu 2025) when targeting HCI/psychology readers.

Scientometrics research on citation determinants (M series, 7 items) backs these observations with general findings. Systematic reviews show higher citation rates than other research designs (Royle et al. 2013: 6.6 per year on average; the journal’s JIF alone explains 59% of variance), and the main predictors of citations are journal prestige, number of authors, international co-authorship, and the volume and recency of references (the 28 factors of Tahamtan et al. 2016; Kousha & Thelwall 2023). Reference-list length correlates positively with citation counts (Mammola et al. 2020), so the review format itself, citing literature thickly, has a structure prone to being cited. Meanwhile, the current consensus on the open-access citation advantage is “contradictory and indeterminate” (the 134-paper review of Langham-Putrow et al. 2021).

Question 4: Verification Results for the Gap Areas

The six gap candidates raised from the previous corpus were verified against the existence of prior reviews and the growth of primary studies.

To state the conclusion first: one is no longer a gap (X04), two are strongly grounded gaps (X02, X05), and three are conditional gaps (X01, X03, X06).

X04 (a meta-analysis of the label effect) has already been filled. De Rooij 2025 published a meta-analysis of 35 experiments and 191 effect sizes from 2017–2024 in Psychology of Aesthetics, Creativity, and the Arts (limited to visual art). What remains is only extension to music and writing, culture-level moderators, and updates covering post-2024 work, so the promise of this angle has declined. Unless gap candidates are verified to this standard before writing, one discovers the prior review only after having written.

X02 (connecting computational-creativity evaluation frameworks to generative AI) is the most strongly grounded. The nearest review, Franceschelli & Musolesi 2024 (ACM Computing Surveys), builds a correspondence table between Boden’s three criteria and machine-learning methods, yet itself places the connection to output evaluation out of scope (explicitly stating that the person/process/press dimensions are deprioritized). The evaluation tutorial of Lamb et al. 2018 stops before LLMs. Meanwhile, primary studies where “computational creativity,” “evaluation,” and “large language model” co-occur grew 15.8-fold from 20 in 2022 to 317 in 2025, and only 3 papers treat Boden, SPECS, and the Lovelace test together with LLMs. Between the stably cited body of theory (over 3,850 cumulative citations) and the surging body of empirical work, no review exists that bundles the two together.

X05 (the offense-defense contest of artist-protection technology) is also strongly grounded. The conflict between the Glaze/Nightshade line of protection and circumvention attacks (Hönig et al.) is technically unresolved, and related primary studies have grown from 2 in 2023 to 18 in 2026. Existing surveys center on IP protection in general or disinformation countermeasures, and no comprehensive review specialized in the style-protection contest could be confirmed across multiple search terms. The literature is scattered across three fields (security, law, and art practice), so an integrative review stands to collect citations from all three.

X06 (the art labor market) has the largest volume of primary studies (604 items) but requires differentiation. A systematic review of the labor market in general already exists, and without a clear specialization in art occupations the work would overlap. The demand for organizing contradictory empirical results (short-term panels find no effect; longitudinal studies report income declines) and the position of being cited by policy documents are its strengths.

X01 (a historical integration from the 1960s to the post-LLM era) has high disciplinary significance but a middling citation forecast. The situation in which art history (Taylor, Klütsch), CS (Franceschelli & Musolesi), and HCI do not cite each other means the absence of a nodal point. However, primary studies attempting the bridge are themselves few (fewer than 10 in related searches), so the citation population is thin.

X03 (LLM creative coding) is premature. Primary studies are growing rapidly, from 1 in 2022 to 57 in 2025, but the absolute number is small and the review target is insufficient. Its promise rises in two to three years.

Synthesis: Ranking the Angles Most Likely to Attract Citations

Integrating the four questions, the angles are ranked by “citation demand (who will cite it),” “competition (existence of prior reviews),” and “growth (increase in primary studies).”

Rank 1: a review connecting computational-creativity evaluation frameworks to the generative-AI era (X02). Three reasons overlap. It can stand at the nodal point that re-cites the stably cited body of theory (over 3,850 cumulative citations) in one stroke; the primary studies it would connect are growing at a near-fastest 15.8-fold; and the nearest review acknowledges the absence of the connection, so there is no competition. The McCormack contrast seen in Question 1 (6 per year when confined to the older frame, 21.9 per year when connected to the current controversy) also supports the advantage of the connecting type. The intended readership spans three communities: computational creativity, HCI, and empirical aesthetics.

Rank 2: a review bundling human-AI co-creation along the axis of evaluation. Theme growth is the fastest of all candidates (+607%), and citation demand is the largest. However, the HCI-side systematic reviews (Hu et al. 2025, Tsao et al. 2025) are already growing fast, so on the same ground (interaction classification, practitioner perceptions) one would be a latecomer. The axis that does not overlap with the competition is evaluation (how to measure the outcomes and process of co-creation), which can partially merge with the Rank 1 angle.

Rank 3: a review of the artist-protection offense-defense contest (X05). The certainty of absent competition is high, and it would be cited from three fields (security, law, and art practice). Because the absolute number of primary studies is still small (44 items), bundling it with the copyright debate (+160% growth) as “the technological and legal contest over style” broadens the population.

Rank 4: an empirical review of generative AI and the art labor market (X06). The population of primary studies is the largest (604 items), and organizing contradictory empirical results is a clear contribution. It requires differentiation by specializing in art occupations and the ability to handle the conventions of both economics and HCI.

Rank 5: a historical integration review (X01). Feasibility is highest because the previous 94-item corpus serves directly as material, and the significance of bridging the field’s fragmentation is also large. However, the citation population is thin, and the expected citations fall short of Ranks 1 and 2. A structure that absorbs it as the introduction to the Rank 1 angle (the historical genealogy of evaluation frameworks) can capture both significance and citations.

Rank 6: an updated meta-analysis of the label effect (the remainder of X04). Given the existence of De Rooij 2025, it holds only as an update or an extension (music and writing, cultural regions). New experiments are accumulating at a pace of 165 per year, so an update in two to three years is meaningful.

Rank 7: a review of LLM creative coding (X03). The review target is currently insufficient. Keep monitoring, and re-evaluate once the population reaches three digits.

The structural recommendations follow mechanically from Question 3. Taxonomy and an open-problems section are mandatory; organized definitions of the target domain (sorting out the 1960s origin of the term generative art and its semantic shift after generative AI becomes naturally necessary for the Rank 1 and Rank 5 angles); a comparison table of evaluation methods; a published accompanying literature repository; and thick references. The observed pattern is narrative survey + taxonomy for CS venues, and PRISMA-ScR compliance for HCI/psychology venues.

Unverified Items

The corpus carries 13 [要一次検証] (primary-verification-needed) markers, all concerning bibliographic and numerical confirmation.

  1. R05: citation velocity for Romero & Machado 2008 cannot be computed due to an OpenAlex publication-year discrepancy (recorded as 2014)
  2. R29: Semantic Scholar citation count for He et al. 2024 (not retrieved due to rate limit) and the published-version DOI
  3. R42: citation count for the Zhao et al. LLM survey (not registered in OpenAlex; S2 rate limit)
  4. R43: citation count for Salas Espasa & Camacho 2025
  5. G02: the narrowing query to text-to-image “evaluation studies” was not run (technical papers are mixed in)
  6. G05: complementary queries for creative coding (p5.js/Processing) were not run, so the count may be an underestimate
  7. G07: the formal DOI for Żylińska 2020 (confirmed to exist and be reachable as an OA book from OHP)
  8. M07: the main figures of Ho et al. 2016 (authentication wall)
  9. S03/S04/S05: details of structural elements (full text unconfirmed due to ACM DL/ScienceDirect authentication walls)
  10. X05: whether Šarčević et al. 2024 mentions Glaze/Nightshade
  11. X06: the exact bibliography of the general labor-market SR
  12. R08: the published-version DOI for Broad et al. 2021 (whether an MDPI Entropy version exists)
  13. The figure for Glaze’s adoption scale had only a media source (MIT Technology Review) and was excluded from this corpus (to be recorded separately if a primary source can be confirmed)

References

All accessed 2026-08-02. Ids correspond to rows in the corpus source/review/generative-art-review-citation-strategy/papers.md. For the sources of citation counts, see the corpus (OpenAlex / Semantic Scholar, retrieved 2026-08-02).

R Series (Measured Citations of Existing Reviews and Key Literature, Excerpt)

  1. (R01) Galanter, P. 2016. Generative Art Theory. In A Companion to Digital Art. Wiley. https://doi.org/10.1002/9781118475249.ch5
  2. (R02) McCormack, J.; Gifford, T.; Hutchings, P. 2019. Autonomy, Authenticity, Authorship and Intention in Computer Generated Art. EvoMUSART 2019. https://doi.org/10.1007/978-3-030-16667-0_3
  3. (R03) McCormack, J. et al. 2014. Ten Questions Concerning Generative Computer Art. Leonardo 47(2). https://doi.org/10.1162/LEON_a_00533
  4. (R04) McCormack, J.; Cruz Gambardella, C. 2022. Complexity and Aesthetics in Generative and Evolutionary Art. Genetic Programming and Evolvable Machines 23. https://doi.org/10.1007/s10710-022-09429-9
  5. (R05) Romero, J.; Machado, P. (eds.) 2008. The Art of Artificial Evolution. Springer. https://doi.org/10.1007/978-3-540-72877-1
  6. (R06) McCormack, J. 2005. Open Problems in Evolutionary Music and Art. LNCS 3449. https://doi.org/10.1007/978-3-540-32003-6_43
  7. (R07) Mazzone, M.; Elgammal, A. 2019. Art, Creativity, and the Potential of Artificial Intelligence. Arts 8(1). https://doi.org/10.3390/arts8010026
  8. (R08) Broad, T.; Berns, S.; Colton, S.; Grierson, M. 2021. Active Divergence with Generative Deep Learning: A Survey and Taxonomy. arXiv. https://arxiv.org/abs/2107.05599
  9. (R09) Maerten, A.-S.; Soydaner, D. 2023. From Paintbrush to Pixel. arXiv. https://arxiv.org/abs/2302.10913
  10. (R10) Chamberlain, R. et al. 2018. Putting the Art in Artificial. Psychology of Aesthetics, Creativity, and the Arts 12(2). https://doi.org/10.1037/aca0000136
  11. (R11) Bellaiche, L. et al. 2023. Humans versus AI. Cognitive Research: Principles and Implications 8. https://doi.org/10.1186/s41235-023-00499-6
  12. (R12) Hong, J.-W.; Curran, N. 2019. Artificial Intelligence, Artists, and Art. ACM TOMM 15(2s). https://doi.org/10.1145/3326337
  13. (R13) Gangadharbatla, H. 2022. The Role of AI Attribution Knowledge in the Evaluation of Artwork. Empirical Studies of the Arts 40. https://doi.org/10.1177/0276237421994697
  14. (R14) Tsao, J. et al. 2025. Perceptions and Integration of Generative AI in Creative Practices and Industries: A Scoping Review. AI & Society. https://doi.org/10.1007/s00146-025-02667-2
  15. (R15) Lamb, C.; Brown, D. G.; Clarke, C. L. A. 2018. Evaluating Computational Creativity: An Interdisciplinary Tutorial. ACM Computing Surveys 51(2). https://doi.org/10.1145/3167476
  16. (R16) Oppenlaender, J. et al. 2024. Prompting AI Art. International Journal of Human-Computer Interaction. https://doi.org/10.1080/10447318.2024.2431761
  17. (R17) Zhou, E. B.; Lee, D. 2024. Generative Artificial Intelligence, Human Creativity, and Art. PNAS Nexus 3(3). https://doi.org/10.1093/pnasnexus/pgae052
  18. (R18) Hu, X. et al. 2025. Designing Interactions with Generative AI for Art and Creativity: A Systematic Review and Taxonomy. DIS 2025. https://doi.org/10.1145/3715336.3735843
  19. (R19) Naqvi, S. M. et al. 2025. Catalyst for Creativity or a Hollow Trend? CHI 2025. https://doi.org/10.1145/3706598.3713233
  20. (R20) Shelby, R. et al. 2024. Generative AI in Creative Practice. CHI 2024. https://doi.org/10.1145/3613904.3642461
  21. (R21) Takagi, H. 2001. Interactive Evolutionary Computation. Proceedings of the IEEE 89(9). https://doi.org/10.1109/5.949485
  22. (R22) Boden, M. A. 1998. Creativity and Artificial Intelligence. Artificial Intelligence 103. https://doi.org/10.1016/S0004-3702(98)00055-1
  23. (R23) Boden, M. A. 2009. Computer Models of Creativity. AI Magazine 30(3). https://doi.org/10.1609/aimag.v30i3.2254
  24. (R24) Colton, S.; Wiggins, G. A. 2012. Computational Creativity: The Final Frontier? ECAI 2012. https://doi.org/10.3233/978-1-61499-098-7-21
  25. (R25) Jordanous, A. 2012. A Standardised Procedure for Evaluating Creative Systems. Cognitive Computation 4(3). https://doi.org/10.1007/s12559-012-9156-1
  26. (R26) Ritchie, G. 2007. Some Empirical Criteria for Attributing Creativity to a Computer Program. Minds and Machines 17(1). https://doi.org/10.1007/s11023-007-9066-2
  27. (R27) Wiggins, G. A. 2006. A Preliminary Framework for Description, Analysis and Comparison of Creative Systems. Knowledge-Based Systems 19(7). https://doi.org/10.1016/j.knosys.2006.04.009
  28. (R28) Jordanous, A. 2016. Four PPPPerspectives on Computational Creativity. Connection Science 28(2). https://doi.org/10.1080/09540091.2016.1151860
  29. (R29) He, Y. et al. 2024. LLMs Meet Multimodal Generation and Editing: A Survey. arXiv. https://arxiv.org/abs/2405.19334
  30. (R30) Bie, F. et al. 2024. RenAIssance: A Survey Into AI Text-to-Image Generation in the Era of Large Model. IEEE TPAMI. https://doi.org/10.1109/TPAMI.2024.3522305
  31. (R31) Huang, Y. et al. 2025. Diffusion Model-Based Image Editing: A Survey. IEEE TPAMI. https://doi.org/10.1109/TPAMI.2025.3541625
  32. (R32) Chen, H. et al. 2025. Comprehensive Exploration of Diffusion Models in Image Generation: A Survey. Artificial Intelligence Review. https://doi.org/10.1007/s10462-025-11110-3
  33. (R33) Lovato, J. et al. 2024. Foregrounding Artist Opinions. AIES 2024. https://doi.org/10.1609/aies.v7i1.31691
  34. (R34) Lemley, M. A. 2024. How Generative AI Turns Copyright Law Upside Down. Science and Technology Law Review 25(2). https://doi.org/10.52214/stlr.v25i2.12761
  35. (R35) Buick, A. 2024. Copyright and AI Training Data—Transparency to the Rescue? Journal of Intellectual Property Law & Practice. https://doi.org/10.1093/jiplp/jpae102
  36. (R36) Yang, L. et al. 2023. Diffusion Models: A Comprehensive Survey of Methods and Applications. ACM Computing Surveys 56(4). https://doi.org/10.1145/3626235
  37. (R37) Wu, S. et al. 2023. A Comprehensive Survey of AI-Generated Content (AIGC). arXiv. https://doi.org/10.48550/arxiv.2303.04226
  38. (R38) Lim, W. M. et al. 2023. The Power of Generative AI. Future Internet 15(8). https://doi.org/10.3390/fi15080260
  39. (R39) Franceschelli, G.; Musolesi, M. 2024. Creativity and Machine Learning: A Survey. ACM Computing Surveys 56(11). https://doi.org/10.1145/3664595
  40. (R40) Wang, B.; Chen, Q.; Wang, Z. 2025. Diffusion-Based Visual Art Creation: A Survey and New Perspectives. ACM Computing Surveys. https://doi.org/10.1145/3728459
  41. (R41) De Rooij, A. 2025. Bias Against Artificial Intelligence in Visual Art: A Meta-Analysis. Psychology of Aesthetics, Creativity, and the Arts. https://doi.org/10.1037/aca0000833
  42. (R42) Zhao, W. X. et al. 2023. A Survey of Large Language Models. arXiv. https://arxiv.org/abs/2303.18223
  43. (R43) Salas Espasa, C.; Camacho, M. 2025. From Aura to Semi-Aura: Reframing Authenticity in AI-Generated Art. AI & Society 40. https://doi.org/10.1007/s00146-025-02361-3

S Series (Surveys Whose Structural Elements Were Confirmed On-Site, Excluding Overlap with the R Series)

  1. (S03) Saxena, D.; Cao, J. 2021. Generative Adversarial Networks (GANs): Challenges, Solutions, and Future Directions. ACM Computing Surveys 54(3). https://doi.org/10.1145/3446374
  2. (S04) Chang, Y. et al. 2024. A Survey on Evaluation of Large Language Models. ACM TIST. https://doi.org/10.1145/3641289
  3. (S05) Frolov, S. et al. 2021. Adversarial Text-to-Image Synthesis: A Review. Neural Networks 144. https://doi.org/10.1016/j.neunet.2021.07.019
  4. (S06) Agnese, J. et al. 2020. A Survey and Taxonomy of Adversarial Neural Networks for Text-to-Image Synthesis. WIREs Data Mining and Knowledge Discovery. https://doi.org/10.1002/widm.1345

M Series (Meta-Research on Citation Determinants)

  1. (M01) Tahamtan, I.; Safipour Afshar, A.; Ahamdzadeh, K. 2016. Factors Affecting Number of Citations: A Comprehensive Review of the Literature. Scientometrics 107(3). https://doi.org/10.1007/s11192-016-1889-2
  2. (M02) Kousha, K.; Thelwall, M. 2023. Factors Associating with or Predicting More Cited or Higher Quality Journal Articles. JASIST 75(3). https://doi.org/10.1002/asi.24810
  3. (M03) Langham-Putrow, A.; Bakker, C.; Riegelman, A. 2021. Is the Open Access Citation Advantage Real? PLOS ONE. https://doi.org/10.1371/journal.pone.0253129
  4. (M04) Blümel, C.; Schniedermann, A. 2020. Studying Review Articles in Scientometrics and Beyond. Scientometrics 124(1). https://doi.org/10.1007/s11192-020-03431-7
  5. (M05) Royle, P. et al. 2013. Bibliometrics of Systematic Reviews. Systematic Reviews 2:74. https://doi.org/10.1186/2046-4053-2-74
  6. (M06) Mammola, S. et al. 2020. Impact of the Reference List Features on the Number of Citations. Scientometrics. https://doi.org/10.1007/s11192-020-03759-0
  7. (M07) Ho, M. H.-C.; Liu, J. S.; Chang, K. C.-T. 2016. To Include or Not: The Role of Review Papers in Citation-Based Analysis. Scientometrics. https://doi.org/10.1007/s11192-016-2158-0

Adjacent Literature Confirmed in the X Series (Excluding Overlap with the Above)

  1. Hönig, R.; Rando, J.; Carlini, N.; Tramèr, F. 2024. Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI. ICLR 2025. https://doi.org/10.48550/arXiv.2406.12027
  2. Šarčević, T. et al. 2024. U Can’t Gen This? A Survey of IP Protection Methods for Data in Generative AI. arXiv. https://arxiv.org/abs/2406.15386
  3. Deng, J. et al. 2024. A Survey of Defenses Against AI-Generated Visual Media. arXiv. https://arxiv.org/abs/2407.10575
  4. Makridis, C. 2026. The Labor Market Effect of Generative Artificial Intelligence on Artists. Journal of Cultural Economics. https://doi.org/10.1007/s10824-026-09575-3
  5. Żylińska, J. 2020. AI Art: Machine Visions and Warped Dreams. Open Humanities Press. https://openhumanitiespress.org/books/titles/ai-art/
  6. Chesterman, S. 2024. Good Models Borrow, Great Models Steal: Intellectual Property Rights and Generative AI. Policy and Society. https://doi.org/10.1093/polsoc/puae006
  7. Doshi, A. R.; Hauser, O. P. 2024. Generative AI Enhances Individual Creativity but Reduces the Collective Diversity of Novel Content. Science Advances. https://doi.org/10.1126/sciadv.adn5290
  8. Chung, J. J. Y. et al. 2022. TaleBrush: Sketching Stories with Generative Pretrained Language Models. CHI 2022. https://doi.org/10.1145/3491102.3501819
  9. Angert, T. et al. 2023. Spellburst: A Node-based Interface for Exploratory Creative Coding with Natural Language Prompts. UIST 2023. https://doi.org/10.1145/3586183.3606719

(For the underlying data of theme-level publication counts, see the list of OpenAlex API URLs in the corpus Provenance section)

← All Notes · Home