GX & Decarbonization Research

Kaiwei Jia, Wenhua Wang

An empirical study using double machine learning (DML) to causally test whether AI adoption mitigates or exacerbates supply chain carbon footprints. It offers causal evidence on the ambiguous environmental effects of AI, relevant to Scope 3…

Fabian Dablander, Simon Wimmer, Jonas M B Haslbeck

First comprehensive LLM-based analysis of 18,785 Guardian climate articles (2010–2025). Coverage rose sharply after 2018 and stayed elevated; causes and mitigation focus on fossil fuels and renewables, while agriculture, overconsumption, ca…

Shen S.

This paper applies LSTM and Transformer deep-learning models to forecast the development trend of the green finance market. By combining long-term temporal dependency with attention mechanisms, it aims to predict future movements of green f…

Chen W.

This paper proposes federated learning analytics to verify and assure the authenticity of ESG data within digital apparel ecosystems. From a platform governance perspective, it addresses how to secure trust in distributed supply-chain ESG i…

Anis Omri, Fadhila Hamza

Using panel data from 23 emerging economies (2015–2023), this study shows AI advances the renewable energy transition partly through green technological innovation. Cybersecurity strengthens AI's effect on both innovation and renewables, wh…

Aisu Mustafaevna Bodur, Victor Sergeevich Tokarev, Alexander Sergeevich Nevzorov

This paper reviews digital technologies for reducing agriculture's carbon footprint and supporting sustainable development. It surveys global and Russian GHG emissions and examines smart greenhouses, smart farming, IoT, AI, and big data. Th…

Jinglong Chen, Qinyi Gu, Wencheng Lin +1

This study applies a large language model to conduct multitopic analysis of how China's low-carbon policies diffuse across regions and time. It demonstrates an AI-driven method for mapping policy theme propagation from policy text. As an ea…

Anonymous 2 Anonymous2

A zero-shot multi-scope LCA framework estimates ML workload footprints without executing models, using DAG traversal and topological depth extraction. It unifies operational carbon, Scope 3 embodied hardware carbon, network transmission car…

Funny Chibwe, Tendai James Mapanga

This study examines expert perspectives on AI-enabled ESG carbon accounting and reporting in Zimbabwe. It highlights AI's potential to automate carbon measurement and disclosure while identifying data infrastructure and institutional gaps a…

Danqing Lin, Dan He, Yu Zhao

This study introduces a multi-agent LLM framework that extracts ESG and financial indicators directly from banks' annual reports, using category-specific prompts to handle synonyms, scales and unit inconsistencies. Applied to hundreds of re…

Felix Maximilian Kania

This study matches facility-level GHG emissions from the US GHGRP self-reports against Climate TRACE satellite estimates and uses machine learning and Bayesian inference to identify drivers of discrepancies. Facility-specific effects explai…

Elsa Manora Simamora

This qualitative single-case study examines how Bank Rakyat Indonesia, the country's largest state-owned bank, integrates ESG principles with AI capabilities in its digital transformation strategy, analysing 47 annual, sustainability and re…

Randy Adhiputra, A. A. S. Gunawan, Ferry Hadiyanto

Using two-way fixed-effects panel regression and tuned tree-based ML ensembles on FTSE Developed ESG Index firms (2016-2025), this study finds ESG positively associated with Tobin's Q, marginally with ROA, and not with ROE, with effects con…

S. B., Vijaya Maheswari Velala, Ahmad Azhar +2

Using survey data from Indian commercial banks, this study examines how corporate governance, internal audit quality, and AI adoption jointly affect sustainable financial reporting quality. Internal audit quality and governance are positive…

B. T, Sabarigiri Selvaraj, Tamil Priyan R. K. B +3

This study evaluates sustainable concrete using ceramic waste and powder to partially replace coarse aggregates (0-100%) and cement (0-35%). At 35% cement replacement, 28-day compressive strength dropped only ~2.5% while embodied CO2 emissi…

Xiutian Shi, Yanan Chen, Shuai Liu

This paper uses a game-theoretic model to examine how AI-enabled ESG management reshapes operational decisions in a supplier-retailer supply chain with responsibility investment, information distortion, and consumer privacy concerns. The su…

Bayu Tri Cahya, Dwi Putri Restuti, A Okfitasari +1

This chapter examines how AI-driven tools—machine learning, digital twins, remote sensing, and real-time MRV—can transform carbon accounting beyond retrospective emissions reporting. Grounded in the GHG Protocol and Legitimacy/Stakeholder T…

Kenedy Alva, Brenda Morales, Paola Ovando

This paper tests whether the linguistic content of forest project descriptions in major voluntary carbon registries predicts credit issuance and retirement. Using NLP (lemmatisation, bigrams, semantic families, domain classification) in log…

Yo-Der Huang, Unaffiliated, Chun-Che Huang +1

This study addresses emission trading (ET) as a multi-objective matching problem between polluters and green entities. It proposes EMOBET, an evolutionary multi-objective optimization approach that simultaneously minimizes buyer cost, maxim…

Flávia Mendes de Almeida Collaço, Felipe Moura Oliveira, Thiago Costa Holanda +1

Analyzing 286 Brazilian newspaper articles (G1, Estadão; Jan 2020–Jun 2023) with BERTopic and supervised LDA, this study finds green hydrogen coverage prioritized financing, investment, production costs, regulation, and exports. Social acce…