The exact text we saved

This is the version exactly as it was saved, character for character. The SHA-256 hash below is a 64-character code worked out from that text. Run the same text through SHA-256 and you get the same code; change a single letter and the code comes out completely different. Each version also stores the previous version’s hash, which chains them together so no version can be swapped out quietly.

TRUVACE RECORD VERSION
record: TRV-2026-2762
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-12T17:20:26.617806Z
status: published
lens: g_space
sector: education
headline: Enhancing students’ Creativity with Generative Artificial Intelligence: A Systematic Review and Meta-Analysis
dek: The rapid advancement of generative artificial intelligence (GAI) is reshaping educational practices and increasingly influencing students’ learning outcomes. Although numerous studies have examined the impact of GAI on students’ creativity, their findings have remained inconsistent. To clarify this relationship and identify key moderation factors, we conducted a systematic review and meta-analysis of 19 empirical studies following PRISMA guidelines. The results revealed a positive and statistically significant…
gain_title: A systematic review and meta-analysis of 19 empirical studies found generative AI had a positive, statistically significant effect on students' creativity (g = 0.908), seen across research designs and educational settings.
problem_title: (none)
trace_subject: (none)
gain_reading: A systematic review and meta-analysis of 19 empirical studies found generative AI had a positive, statistically significant effect on students' creativity (g = 0.908), seen across research designs and educational settings.
gain_evidence: positive and statistically significant overall effect of GAI on students’ creativity (g = 0.908)
problem_reading: (none)
problem_evidence: (none)
quick_read: A systematic review and meta-analysis following PRISMA guidelines pooled 19 empirical studies on generative AI and students' creativity. It found a positive, statistically significant overall effect (g = 0.908), observed across diverse research designs and educational contexts.

The authors say earlier findings on the question had been inconsistent, so the pooled estimate aims to settle it. Three factors significantly shaped the size of the effect: the role the AI played, how students interacted with it, and how long the intervention lasted.

Output modality, task difficulty and research design did not significantly change the effect. The authors present the results as support for using generative AI as an educational support tool in teaching.
limitation: 
tag: Evidence-backed gain
key_points: The review pooled 19 empirical studies on generative AI and student creativity, following PRISMA guidelines. | The overall pooled effect was positive and statistically significant (g = 0.908). | The effect appeared across diverse research designs and educational contexts. | The role given to the AI significantly moderated the effect. | The mode of student-AI interaction significantly moderated the effect. | Intervention duration also significantly moderated the effect. | Output modality, task difficulty and research design showed no significant moderating effect.
rundown: The researchers set out to resolve inconsistent findings from earlier studies of how generative AI affects students' creativity. They ran a systematic review and meta-analysis of 19 empirical studies, following PRISMA reporting guidelines.

The pooled result was a positive, statistically significant effect (g = 0.908), and it held across diverse research designs and educational contexts.

Moderation analyses found that the role adopted by the AI, the mode of student-AI interaction and the duration of the intervention significantly shaped the effect. The type of output the AI generated, task difficulty and research design did not.
sources:
- peer_reviewed | Creativity Research Journal | https://doi.org/10.1080/10400419.2026.2726242 | 2026-09-07
prev: 0000000000000000000000000000000000000000000000000000000000000000

SHA-256 hash

280492bbb8402a67e4719f06ebbb35ac214ef1f4c70c90d21a22a75ea55f7863

Previous version’s hash

0000000000000000000000000000000000000000000000000000000000000000