Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 83 of the 84 references it cites were found in a registry or an index; 17 of the 58 sources its claims cite were checked for support.

From humans to AI: understanding why AI is perceived as the preferred co-creation partner

Open access
Liu, Yuchang, Yang, Yongzhong, Xu, Haoran · 2025 · Frontiers in Psychology

Abstract

With the widespread adoption of generative AI in creative industries, individuals increasingly face a choice between human-human co-creation and human-AI co-creation. Prior comparisons of these modes have largely focused on output quality, efficiency, and user experience, while giving less attention to co-creation intention. Drawing on creativity theory, we argue that perceived novelty and perceived usefulness are the key mechanisms linking co-creator types to co-creation intention, and we test this account across four empirical studies. The results show that, relative to co-creating with humans, co-creating with AI significantly increases participants' perceived novelty and, counterintuitively, perceived usefulness, thereby increasing co-creation intention. Qualitative interviews identify three principal drivers of why AI is regarded as more useful-efficiency, value, and relationship. Furthermore, we find that the need to belong exerts a moderating effect. Overall, this research extends creativity theory to the AI collaboration context, challenges the conventional assumption that "AI offers greater novelty whereas humans offer greater usefulness," and uncovers social-motivational boundary conditions in technology-assisted creative work.

Overview

CoChat Summary

## TL;DR

Across four studies, the paper finds that people are generally more willing to co-create with AI than with humans, because AI increases both perceived novelty and perceived usefulness. It also shows that the usefulness pathway is weaker for people with a higher need to belong.

## Key findings

  • In Study 1 (n = 150), human–AI co-creation increased perceived novelty (M_AI = 5.80 vs. M_human = 4.07; F(1,148) = 60.35, p < 0.001, ηp2 = 0.29), perceived usefulness (M_AI = 5.89 vs. M_human = 5.31; F(1,148) = 13.03, p < 0.001, ηp2 = 0.08), and co-creation intention (M_AI = 6.12 vs. M_human = 5.35; F(1,148) = 23.29, p < 0.001, ηp2 = 0.14).
  • Study 1 mediation: the total effect of co-creator type on co-creation intention was significant (β = 0.769, p < 0.001), the direct effect was not (β = 0.041, p = 0.635), and both indirect effects were significant via perceived novelty (β = 0.292) and perceived usefulness (β = 0.436).
  • In Study 2 (n = 243), the pattern replicated: AI increased perceived novelty (M_AI = 5.55 vs. M_human = 4.60; F(1,241) = 34.92, p < 0.001, ηp2 = 0.13), perceived usefulness (M_AI = 5.88 vs. M_human = 5.72; F(1,241) = 4.52, p = 0.035, ηp2 = 0.02), and co-creation intention (M_AI = 5.96 vs. M_human = 5.71; F(1,241) = 7.05, p = 0.008, ηp2 = 0.028).
  • Study 3 interviews (n = 50) found that 76% (n = 38) preferred co-creating with AI, 92% (n = 46) viewed AI as more novel, and 74% (n = 37) viewed AI as more useful; the main reasons AI seemed useful were efficiency, value, and relationship.
  • In Study 4 (n = 294), the total effect remained significant (β = 0.221, p = 0.005) and the direct effect was not (β = −0.051, p = 0.441); the interaction between co-creator type and need to belong was significant for usefulness (β = −0.466, p < 0.001) but not for novelty (β = −0.193, p = 0.201).

## Methods

  • The paper used four studies: three scenario-based experiments (Studies 1, 2, and 4) and one semi-structured interview study (Study 3), comparing human vs. AI as co-creation partners.
  • Sample sizes were 150, 243, 50, and 294 participants respectively; all experimental participants had prior experience using generative AI and were recruited via Credamo, while Study 3 recruited participants with experience co-creating with both AI and humans.
  • Main analyses included ANOVAs for basic effects, regression analyses, mediation analysis with Hayes’s PROCESS Model 4 (5,000 bootstraps), and moderated mediation with PROCESS Model 7 in Study 4; Study 3 also used word-frequency and thematic analysis.

## Limitations

  • The paper states that its tasks were scenario-based and “still differ from fully authentic content co-creation”; it notes that future work should use field experiments in organizational and platform settings to strengthen external validity.
  • It states that the studies focused on writing, image-generation, and one cognitive-analytical writing task, while paying less attention to more complex operations such as model design and video production.
  • It states that the samples were recruited in China via Credamo, which limits generalizability and raises cultural alternative explanations.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3389/fpsyg.2025.1695532

Cites

83 of 84 references verified1 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Research model.
Figure 1 Research model.CC BY · PubMed Central
Figure 2 Effects of co-creator types (Study 1).
Figure 2 Effects of co-creator types (Study 1).CC BY · PubMed Central
Figure 3 Effects of co-creator types (Study 2).
Figure 3 Effects of co-creator types (Study 2).CC BY · PubMed Central
Figure 4 Drivers of AI’s Usefulness.
Figure 4 Drivers of AI’s Usefulness.CC BY · PubMed Central
Figure 5 Interaction plot of the moderation effect.
Figure 5 Interaction plot of the moderation effect.CC BY · PubMed Central

Verification

Claims the paper makes

12 Valid · 4 Not valid · 21 Not checked · 98 Own results and interpretations

71 more claims · 1 with their cited sources checked

Create a free account to see all 135 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 25
  • It may state two separate claims as one. 42
  • The quote only partly states this claim. 2

Create a free account to see these claims.

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • ConsistentF(1, 148) = 60.35reported p < .001 · recomputed p < .001

    “F(1,148) = 60.35, p < .001”

  • ConsistentF(1, 148) = 13.03reported p < .001 · recomputed p < .001

    “F(1,148) = 13.03, p < .001”

  • ConsistentF(1, 148) = 23.29reported p < .001 · recomputed p < .001

    “F(1,148) = 23.29, p < .001”

  • Consistentt(146) = 5.57reported p < .001 · recomputed p < .001

    “t(146) = 5.57, p < .001”

  • Consistentt(146) = 17.92reported p < .001 · recomputed p < .001

    “t(146) = 17.92, p < .001”

5 of 10 · the rest aren’t loaded

5 more recomputed tests

Create a free account to see every test the Stats check recomputed.

See all tests — free account

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Citations

Cites

Show all 82 references

Research works printed but not linked: 2

  1. Hayes, 2017, Introduction to mediation, moderation, and conditional process analysis: a regression-based approach
  2. Zhang, 2025, 与人共创还是与AI共创?共创主体类型对内容共创意愿的影响研究 [co-creation with humans or AI? The influence of co-creator types on content co-creation intention], J. Psychol. Sci., 48, 210, 10.16719/j.cnki.1671-6981.20250120 · doi:10.16719/j.cnki.1671-6981.20250120

Metadata

Authors
Liu, Y., Yang, Y., & Xu, H.
Affiliations
  • Zhejiang Gongshang University (CN)
  • Sichuan University (CN)
Corresponding author
Xu, Haoran
Venue
Frontiers in Psychology
Published
2025-12-08
Volume
16
Pages
1695532-1695532
ISSN
1664-1078 (ISSN-L)
OpenAlex
W4417121594
Cited by
10
References
82
Citations by year
2026: 10
FWCI
10.47
Type
journal article
Language
en
Open access
Locations
  • Frontiers in Psychology (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Creativity in Education and Neuroscience
  • Physical Sciences › Computer Science › Computer Graphics and Computer-Aided Design › Digital Media and Visual Art
  • Physical Sciences › Computer Science › Human-Computer Interaction › Innovative Human-Technology Interaction
Keywords
human-AI co-creation, generative AI, perceived novelty, perceived usefulness, creativity theory, need to belong

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 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

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: From humans to AI: understanding why AI is perceived as the preferred co-creation partner. https://app.cochat.ai/papers/10.3389/fpsyg.2025.1695532
The paper
Liu, Y., Yang, Y., & Xu, H. (2025). From humans to AI: understanding why AI is perceived as the preferred co-creation partner. Frontiers in Psychology. https://doi.org/10.3389/fpsyg.2025.1695532

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.