AI-assisted teams have a higher probability of producing top-5%-FWCI publications than human-written teams¶

Claim · All claims
Evidence · 1 study · associational r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? AI-assisted publications more often reach the top 5% of FWCI (PLoS 7.34% vs 5.01% pre-GPT and 4.31% post-GPT; Nature 7.40%), and logistic coefficients are positive (2.66; 6.36). → Haoyang Wang 2026

Evidence¶

Haoyang Wang 2026¶

Haoyang Wang, Mingze Zhang, Yi Bu, Star Xing Zhao, Meijun Liu. (2026). Smaller, Younger, and More Impactful: How AI-Assisted Writing Transforms Research Teams. https://arxiv.org/abs/2605.27404 doi:10.48550/arxiv.2605.27404

q2 · i? · associational · r2

Comparison of top-5%-FWCI rates across AI-assisted and human-written groups, with logistic regression (coefficients 2.66 PLoS, 6.36 Nature, p<0.01) and PSM (ATT +3.0% PLoS, +2.7% Nature, p<0.01). The article reports "7.34% of AI-assisted publications achieved the top 5% of FWCI".

"In the PLoS dataset (Fig. 6b), 7.34% of AI-assisted publications achieved the top 5% of FWCI, significantly higher than the 5.01% in the human-written (pre-GPT) group and 4.31% in the human-written (post-GPT) group."

Discussion¶

Related Claims¶

• Research teams using AI-assisted writing have a lower average career age than human-written teams — related
• Teams using AI-assisted writing are smaller-sized, with the association stronger for Nature publications — related
• Most publications show very low AI usage scores, with higher mean AI usage in PLoS than Nature — related
• The negative association between AI-assisted writing and team age is strongest among the youngest teams — related
• AI-assisted teams comprise more junior authors and fewer senior authors — related