Projects

SESG: Integrated Platform for Enhanced Analysis of Environmental, Social, and Governance (ESG) Reports

Improving ESG reporting with generative AI

Funding: IFAB

Enabling technology: Big Data Analytics, LLM

Sustainable Development Goals

SESG aims to develop a platform to analyze ESG data from non-financial reports to identify common practices, soundness, and the reuse of metrics that could alter the ESG index. The use of advanced analytics, NLP, and LLMs simplifies access to ESG reports, enabling comparative analysis and informed decision-making.

National Centre for HPC, Big Data and Quantum Computing (ICSC), a project funded by the European Union – NextGenerationEU – and by Italy’s National Recovery and Resilience Plan (PNRR) – Mission 4, Component 2.

Objectives

The project aims to develop a platform to support the analysis of non-financial transparency declarations of ESG data, to simulate the consistency of indicator data from real-world scenarios. This stems from the need to better understand and develop ESG reports in an accurate and quantifiable manner.

Initial challenge

The importance of ESG indices is necessary for assessing corporate impact on sustainability issues. Organizations are increasingly expected to implement and execute sustainability-related practices and objectives and to account for these efforts transparently. This project responds to the growing complexity and volume of ESG data, which requires more sophisticated tools for analysis and practical interpretation.

Solution

ESG reports will be improved through the creation of a model capable of fully understanding and analyzing all elements that contribute to their creation. This will be achieved through new models linked to generative AI and the use of LLMs, which will enable the development of an interactive dashboard according to the digital twin reference model developed in Spoke 9.

Benefits

The project will provide tools to identify ethical validation and the impact of ESG metrics. ESG data will be provided to banks and financial services to assess a company’s credit risk and analyze its adaptability to climate-related risks, increasing awareness of ESG metrics and evaluating compliance with sustainability requirements.

IFAB’s role

IFAB contributed to the design and implementation of the data infrastructure and artificial intelligence tools needed to automatically analyze non-financial statements and ESG sustainability reports, transforming heterogeneous documentary sources into comparable insights to support the assessment of environmental, social, and governance performance.

Activities covered the entire data lifecycle: from automatic collection through web scraping, to information extraction via a RAG pipeline deployed on the Leonardo HPC cluster, through to comparative evaluation of language models and the definition of an ESG metrics framework. Results are accessible through a web portal with integrated dashboards and a greenwashing analysis module based on explainable AI.

Participation in the project enabled Fondazione IFAB to consolidate expertise at the intersection of artificial intelligence, large-scale document analysis, and sustainable finance.