The rapid expansion of corporate sustainability reporting has increased transparency requirements while raising concerns about greenwashing driven by selective, narrative-based disclosure. This study assesses the credibility of Environmental, Social, and Governance (ESG) communication by comparing corporate sustainability reports with external media coverage for a sample of 204 large firms operating in Central and Eastern Europe in 2023. Using natural language processing techniques, the analysis constructs a Greenwashing Severity Index (GSI) that captures discrepancies between firms’ ESG self-representation and external public narratives. The index combines ESG-specific focus measures, sentiment analysis, TF–IDF-based term weighting, and topic modeling to quantify imbalances in ESG communication. Results indicate moderate but widespread greenwashing across countries, industries, and firm sizes, with substantial heterogeneity linked to differences in regulatory maturity and stakeholder scrutiny. Higher alignment between corporate disclosures and external narratives is observed among larger firms and in sectors subject to stronger public accountability, while finance, aviation, and online commerce exhibit higher greenwashing severity. A propensity score matching analysis further shows that firms with imbalanced emphasis across ESG dimensions display significantly higher GSI values, consistent with strategic disclosure behavior rather than substantive sustainability engagement. Overall, the findings demonstrate that transparency alone is insufficient to ensure credible ESG communication, highlighting the need for EU sustainability governance to move beyond disclosure-based compliance toward digitalized, data-driven monitoring frameworks that systematically integrate external information sources to curb strategic ESG misrepresentation and enhance corporate accountability under evolving regulatory regimes.

The integration of Artificial Intelligence (AI), including Machine Learning (ML) and Natural Language Processing (NLP), into Environmental, Social, and Governance (ESG) frameworks and sustainable finance has gained significant academic and industry attention. This study presents a bibliometric analysis of research at the intersection of AI, ESG, and sustainable finance, using Web of Science data from 2004 to 2025. Our analysis maps the evolution of research trends, identifies key authors, influential publications, and collaboration networks, and explores the intellectual structure of the field. Using co-citation analysis, keyword co-occurrence mapping, and thematic clustering, we reveal how AI-driven methodologies have shaped ESG assessments, sustainable investment strategies, and financial decision-making. Findings highlight a sharp rise in AI applications in ESG post-2015, particularly in ESG risk modelling, climate analytics, and AI-driven reporting. This study provides a comprehensive overview of the evolving academic discourse, identifying emerging research directions and challenges related to data standardization, transparency, and ethical AI applications in sustainability.

This study aims to evaluate the sustainability performance of EU regions through a comprehensive and data-driven Environmental, Social, Governance (ESG) framework, addressing the increasing demand for regional-level analysis in sustainable finance and policy design. Leveraging Partial Least Squares (PLS) regression and cluster analysis, we construct composite ESG indicators that adjust for economic size using GDP normalization and LOESS smoothing. Drawing on panel data from 2010 to 2023 and over 170 indicators, we model the determinants of ESG performance at both the national and regional levels across the EU-27. Time-based ESG trajectories are assessed using Compound Annual Growth Rates (CAGR), capturing resilience to shocks such as the COVID-19 pandemic and geopolitical instability. Our findings reveal clear spatial disparities in ESG performance, highlighting the structural gaps in governance, environmental quality, and social cohesion. The model captures patterns of convergence and divergence across EU regions and identifies common drivers influencing sustainability outcomes. This paper introduces an integrated framework that combines PLS regression, clustering, and time-based trend analysis to assess ESG performance at the subnational level. The originality of this study lies in its multi-layered approach, offering a replicable and scalable model for evaluating sustainability with direct implications for green finance, policy prioritization, and regional development. This study contributes to the literature by applying advanced data-driven techniques to assess ESG dynamics in complex economic systems.

This study investigates how firms in high-emission sectors progress along the low-carbon transition by analysing the joint dynamics of Management Quality (MQ) and Carbon Performance (CP) using probabilistic modelling and explainable machine-learning methods. Digitalisation is conceptualised as the increasing use of data-driven and algorithmic tools in corporate governance, sustainability monitoring, and regulatory oversight, enabling a more granular assessment of corporate transition pathways across multiple time horizons. Using annual Transition Pathway Initiative data for 175 firms over the period 2018–2025, we apply discrete-time Markov chains to capture state persistence and directional mobility in MQ and CP, while Hidden Markov Models uncover latent performance regimes shaping firms’ transition trajectories across three decarbonisation horizons (2028, 2035, and 2050). To enhance interpretability and policy relevance, CatBoost-based feature importance analysis identifies governance, emissions-related, and sector-specific drivers of transitions between states. The results indicate a steady and highly persistent improvement in Management Quality, reflecting cumulative consolidation of governance structures, while Carbon Performance evolves more slowly and heterogeneously, with only moderate convergence emerging toward the 2050 horizon. Latent-regime estimates reveal a gradual shift from volatile, low-performance pathways toward more stable transition regimes over time. From a policy perspective, the findings suggest that governance improvements alone are insufficient to ensure timely emission reductions, highlighting the need for digitally enabled, sector-specific regulatory incentives and enforcement mechanisms targeting realised Carbon Performance.

The rapid development of artificial intelligence (AI) has led to increased concerns about technological substitution and the skills of the future in the labour market. While conventional automation has been commonly associated with substitution in routine tasks and skill-biased technological change, the impact of AI may differ significantly in nature and direction. This paper proposes a meta-analysis of 321 estimates from the empirical evidence on the impact of AI and automation exposure on labour market outcomes. The dataset is constructed from 19 empirical studies examining employment, wage, and skill-related effects across different countries, sectors, and empirical designs. To ensure comparability across heterogeneous regression specifications, all coefficients are transformed into partial correlation coefficients and analysed using a three-level random-effects meta-analytic model. The results indicate that the overall pooled effect of technological exposure on labour market outcomes is small and statistically insignificant. However, substantial heterogeneity exists across studies. While some papers report negative employment effects associated with automation and robot adoption, others document positive outcomes related to productivity gains, wage increases, or skill upgrading. Outcome-specific analyses for employment, wages, and skill demand also show no statistically significant average effects. Moderator analysis further reveals no systematic differences between AI-specific measures and broader automation indicators. Overall, the findings suggest that the labour market effects of artificial intelligence and automation are highly context dependent. Rather than producing a uniform pattern of job displacement or skill-biased technological change, technological adoption appears to generate a set of mechanisms that vary across sectors, occupations, and institutional settings.