Buyer Artificial Intelligence-Enabled Environmental Governance and Supplier Environmental Controversies: An Organizational Information Processing and Signaling
Yongchao Martin Ma, Xinya Guan
cs.CY · cs.AI · stat.AP
Quick verdict
Claims buyer AI-enabled governance curbs supplier controversies; novel 10-K text measure, but no supplier FE and disclosure-only evidence temper causality.
Summary
This empirical panel study asks whether overseas suppliers' exposure to U.S.-listed buyers' AI-enabled environmental governance reduces suppliers' media-reported environmental controversies. Drawing on organizational information processing theory and signaling theory, the authors construct a sentence-level measure counting 10-K sentences (Items 1 and 7) where AI, environmental, and governance terms co-occur, aggregate it to suppliers weighted by relationship duration, and estimate PPML models with supplier-industry, supplier-country, and year fixed effects on 5,620 supplier-year observations covering 2,505 suppliers in 41 countries (2020-2024). Exposure is negatively associated with next-year controversies (β = -0.282, p < .001), more strongly so where supplier-country AI readiness (β = -0.011) and regulatory quality (β = -0.014) are higher. Robustness spans alternative measures, subsamples, negative binomial estimation, and an instrumental-variable design (KP Wald F = 26.09), though the full-text variant attenuates to marginal significance and the 662-observation positive-controversies subsample shows weaker interactions. The authors acknowledge the main limitations: a correlational design without supplier fixed effects, disclosure-based measurement that may capture rhetoric rather than practice, and untested perceptual mechanisms.
Review complete
Assessment at a glance
5
risk flags identified
● high 0
● medium 3
● low 2
5 strengths · 5 weaknesses · 5 author questions
AI text signal = Binoculars zero-shot detector score over 20 sampled passages (Fast-DetectGPT co-reported) — absence does not prove human authorship, and a flag does not imply misconduct.
Levels reflect verification complexity — not a quality score. Evolve never rates your paper.
Key claims under review
- Suppliers' exposure to buyer AI-enabled environmental governance is negatively associated with supplier environmental controversies the following year (Model 2: β = -0.282, p < .001).
- The negative association strengthens when the supplier's home country has higher AI readiness (Model 3 interaction: β = -0.011, p < .05).
- The negative association strengthens when the supplier's home country has higher regulatory quality (Model 4 interaction: β = -0.014, p < .01).
- Buyer AI-enabled environmental governance generates cross-border governance spillovers operating through enhanced information processing and deterrence signaling to suppliers.
Strengths
Genuinely novel and carefully constructed text measure
Section 3.2.2 requires AI, environmental, and governance lexicon terms to co-occur within one sentence (scaled per 1,000 sentences), a sensible guard against false positives from standalone AI or ESG chatter; terms were manually validated against the 10-K corpus, and ambiguous acronyms were excluded.
Econometrics matched to the outcome
PPML with supplier-industry, supplier-country, and year fixed effects, one-year lags, 1/99 winsorization, and supplier-clustered SEs (Section 3.3) is appropriate for the zero-inflated count DV (mean 0.796, SD 5.804, max 202; Table 2); findings survive negative binomial re-estimation (B7).
Above-norm robustness battery with honest disclosure
Controls target the obvious confound channel
IV design with full diagnostics
Weaknesses
No supplier fixed effects in the baseline
Disclosure-based exposure measure may capture rhetoric
Mechanism conflation: deterrence vs. visibility suppression
Unexplained IV magnitude and contested exclusion restriction
Moderator redundancy and reporting slips
Full Review Report
$1.99 · one-timeUnlock the complete preliminary analysis Evolve produced for this paper.
Report preview
Scope & Claims
The paper asks whether foreign suppliers' exposure to U.S.-listed buyers' AI-enabled environmental governance reduces suppliers' media-reported environmental controversies. Grounded in organizational information processing theory (OIPT) and signaling theory, it tests H1 (a direct negative association), H2 (reinforcement by supplier-country AI readiness), and H3 (reinforcement by supplier-country regulatory quality). The evidence is a 2020-2024 unbalanced panel of 5,620 supplier-year observations covering 2,505 first-tier foreign suppliers of non-financial U.S. buyers across 41 countries, with tax-haven suppliers excluded (Section 3.1). All three hypotheses are reported as supported, and a sentence-level triple-co-occurrence text measure of buyer AI-enabled environmental governance is offered as a reusable methodological contribution.
The remaining sections cover Methodology Assessment, Statistical Evidence, Results & Reproducibility, Limitations & Risks, and Verdict & Recommendations.
- All 5 strengths & 5 weaknesses, analyzed in detail
- 5 risk flags with severity, section & reviewer questions
- Methods & statistical evidence assessment
- 5 questions a peer reviewer would ask
- Full six-dimension evaluation report & recommendations
One-time purchase · this paper only
One-time purchase, this paper only. AI-generated preliminary analysis — not peer review.
Frequently asked questions
What did this study find about AI and supplier environmental controversies?
It found that when a supplier's U.S.-listed buyers disclose more AI-enabled environmental governance in their SEC 10-K filings, that supplier tends to have fewer media-reported environmental controversies the following year. The association is stronger in supplier countries with higher AI readiness and better regulatory quality, based on 5,620 supplier-year observations covering 2,505 suppliers in 41 countries from 2020 to 2024.
How did the authors measure buyer AI-enabled environmental governance?
They counted sentences in buyers' 10-K filings (Items 1 and 7) that simultaneously contain an AI term (e.g., 'machine learning'), an environmental term (e.g., 'emissions'), and a governance term (e.g., 'monitoring'), scaled per 1,000 sentences. Buyer scores were then aggregated to each supplier, weighted by buyer-supplier relationship duration.
Is the study's finding causal?
No. The authors themselves state it is primarily correlational. Predictors are lagged one year and an instrumental-variable design is used, but there are no supplier fixed effects, the exposure measure is based on disclosures rather than actual practice, and the IV estimate is several times larger than the baseline, so causal interpretation remains uncertain.
Is the data or code for this paper publicly available?
The reviewed arXiv text contains no code or data availability statement. The construction relies heavily on commercial/licensed sources (FactSet supply-chain data, RepRisk, LSEG Asset4, Oxford Insights), so independent replication is difficult; only the WGI and World Bank inputs are public.
About this review
This is an AI-generated preliminary review by Evolve.Pub, produced from the openly available text of the paper. It is not peer review and does not represent the authors' views. The canonical record of the paper is at the publisher link above.
- The reviewed text contains no appendices: the sentence lexicons (Appendix A) and robustness tables B1-B7 are referenced but not included, so their exact contents could not be checked.
- The first-page abstract and equations (1)-(3) are garbled by font-encoding artifacts in extraction, but the same abstract reappears intact in the body and all quantities are recoverable from surrounding prose and tables.
- Figures 1-3 are referenced but their graphical content is not present in the extracted text; the moderation plots were assessed only through the reported interaction coefficients.
Go deeper on this paper
Wrote a paper? Stress-test it before reviewers do.
Upload your manuscript for the same structured scrutiny — reasoning, statistics, evidence, figures, and reproducibility.