Please rotate your device to landscape mode to view the charts.

Green innovation optimization for climate change ESG business readiness: role of generative AI in BRICS countries

Journal: European Financial Management

Date: 2025

Author: Noman Arshed; Yassine Bakkar; Marco De Sisto; Mubbasher Munir; Shajara Ul-Durar

Abstract:
Climate change introduces new challenges for businesses which require them to find ways to be resilient. Green innovations contribute to boost Environmental, Social, and Governance (ESG)-readiness leading to just transition without optimization. This study estimates the nonlinear effect of environmental innovation in ESG-readiness against climate change while allowing for the moderating role of citations from regenerative AI-research. We use BRICS countries to conduct the analyses with a machine learning based Panel-QARDL. We find that green innovations trace an inverted U-shaped effect and generative AI shifts this relationship upwards. Findings highlight the role of regenerative AI in boosting green innovation performance.

Link: Google Scholar

Background and Context

Research Focus

This study examines how green innovation affects businesses' Environmental, Social, and Governance (ESG) readiness in BRICS countries (Brazil, Russia, India, China, South Africa).

Generative AI Role

The research investigates whether generative AI research can enhance the effectiveness of environmental patents in building climate change resilience.

Methodology

Using data from 2000-2023, researchers applied a machine learning-based Panel Quantile ARDL model to analyze nonlinear relationships between variables.

Green Innovation Shows Inverted U-Shaped Effect on ESG Readiness

Optimal Level Environmental Patents ESG Readiness Positive Returns Diminishing Returns
  • Environmental patents initially boost ESG readiness as firms adopt sustainable practices and improve operational efficiency.
  • Beyond an optimal threshold (around 12 patents), additional green innovation shows diminishing returns on ESG performance.
  • This inverted U-shape suggests businesses must strategically balance innovation investment to maximize climate resilience benefits.

Generative AI Citations Amplify Green Innovation's Positive Effects

  • The chart shows how different levels of generative AI research citations shift the innovation-ESG relationship upward.
  • Higher AI citation levels enable businesses to achieve greater ESG readiness from the same level of environmental patents.
  • This moderating effect demonstrates AI's potential to optimize and enhance green innovation performance in emerging economies.

Long-Run Coefficients Show Consistent Effects Across ESG Distribution

  • Panel Quantile ARDL estimates reveal how effects vary across low (25th), medium (50th), and high (75th) ESG percentiles.
  • Environmental patents consistently show positive linear effects (0.035-0.054) but negative quadratic effects (-0.001 to -0.002).
  • The AI moderation effect is strongest at lower quantiles (0.002) suggesting AI helps most for less ESG-ready businesses.

Digital Infrastructure and Investment Significantly Boost ESG Readiness

Internet Access +0.1% per 1% increase Enables knowledge sharing Firm Investment +0.5% per 1% increase Builds resilient infrastructure → ESG Readiness ←
  • A 1% increase in internet access leads to a 0.1% increase in ESG readiness through enhanced knowledge sharing.
  • Firm investment shows a stronger effect, with each 1% increase boosting ESG readiness by approximately 0.5%.
  • These control variables highlight the importance of digital and physical infrastructure for climate change preparedness.

Country-Level Predictions Reveal Divergent ESG Trajectories in BRICS

  • The model predicts ESG readiness trends differently across BRICS nations based on their innovation and AI profiles.
  • India and South Africa show overprediction, suggesting their ESG readiness may increase relative to current trajectories.
  • Brazil, China, and Russia show underprediction, indicating potential ESG readiness decline without policy intervention.

Contribution and Implications

  • First study to empirically demonstrate generative AI's moderating role on green innovation and ESG readiness relationships.
  • Provides evidence for optimal green innovation investment levels, helping firms avoid diminishing returns on sustainability spending.
  • Offers policymakers guidance on supporting AI research funding specifically targeted at climate change adaptation in emerging economies.
  • Machine learning methodology enables deployment-ready models for real-time ESG monitoring dashboards in business decision-making contexts.
  • Highlights importance of digital infrastructure and capital investment as complementary factors for enhancing climate change resilience.

Data Sources

  • Finding 1 (Inverted U-Shape): Conceptual visualization based on coefficient signs from Table 6 (EPAT positive, EPAT² negative).
  • Finding 2 (AI Moderation): Projection curves derived from Table 6 coefficients and Figure 10 quadratic fit methodology from the article.
  • Finding 3 (Quantile Coefficients): Data directly from Table 6 showing Panel Quantile ARDL long-run estimates at 25th, 50th, and 75th percentiles.
  • Finding 4 (Control Variables): Coefficient values from Table 6 for Internet Access (INTE) and Firm Investment (GFCF) variables.
  • Finding 5 (Country Predictions): Visualization based on Figure 9 description of actual versus predicted ESG values across BRICS countries.