Related Experiment Video
Updated: Jul 9, 2026
03:14
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Assessing corporate sustainability with large language models: evidence from Europe
Kerstin Forster1,2, Lucas Keil3,4, Victor Wagner1,4,5
1LMU Munich, Munich, Germany.
Nature Communications
|July 7, 2026
Summary
Companies are vital for global sustainability, but tracking their environmental, social, and governance (ESG) progress is challenging. Our machine learning framework analyzes corporate reports to assess ESG transparency and performance, revealing key insights.
Area of Science:
- Corporate Sustainability
- Machine Learning Applications
- Financial Reporting Analysis
Background:
- Global sustainability goals necessitate corporate contributions, yet comprehensive data on environmental, social, and governance (ESG) performance is scarce.
- Existing methods for ESG data extraction are often manual and lack systematic scalability.
- Limited evidence exists on the actual progress and transparency of corporate ESG initiatives.
Purpose of the Study:
- To develop and apply a machine learning framework for systematically extracting ESG indicators from corporate reports.
- To construct a large-scale dataset of corporate ESG observations for European firms.
- To assess ESG transparency and performance trends, evaluating alignment with European Sustainability Reporting Standards (ESRS).
Main Methods:
- Development of a machine learning framework for automated ESG indicator extraction from annual and sustainability reports.
- Analysis of reports from 600 large European firms spanning 2014-2023, generating 2.9 million ESG data points.
- Evaluation of ESG transparency based on ESRS-aligned disclosures and ESG performance using extracted numerical data.
Main Results:
- A significant ESG transparency gap was identified, with top-rated firms disclosing 22% more indicators than bottom-rated firms, though this gap is decreasing.
- Corporate ESG performance shows mixed trends: social indicators are largely stagnant (except gender equality), while environmental indicators exhibit moderate improvement.
- Reported Scope 3 emissions have increased substantially, primarily due to enhanced disclosure practices.
Conclusions:
- The developed open-source machine learning framework provides a systematic method for tracking corporate ESG efforts and transparency.
- Findings highlight the need for improved and standardized ESG reporting to accurately reflect corporate sustainability performance.
- The study underscores the evolving landscape of corporate sustainability disclosure and performance measurement.
Related Concept Videos
01:27Language and Cognition
Language serves as a bridge between ideas and communication, influencing how individuals perceive and interact with the world. Psychologists have long debated whether language shapes thought or vice versa. This discussion gained grip with Edward Sapir and Benjamin Lee Whorf in the 1940s, who proposed that language determines thought, a concept known as linguistic determinism. They suggested that the vocabulary and structure of a language influence how its speakers think and perceive reality.
02:16Stereotype Content Model
The Stereotype Content Model (SCM) was first proposed by Susan Fiske and her colleagues (Fiske, Cuddy, Glick & Xu, 2002; see also Fiske, 2012 and Fiske, 2017). The SCM specifies that when someone encounters a new group, they will stereotype them based on two metrics: warmth—or that group’s perceived intent, and how likely they are to provide help or inflict harm—and competence—or their ability to carry out that objective. Depending on the warmth-competence categorization, a person will feel...
01:30Typical Model Studies
Fluid mechanics model studies often utilize scaled-down systems to predict fluid behavior in full-scale environments, such as river flows, dam spillways, and structures interacting with open surfaces. Maintaining Froude number similarity in river models is crucial, as it replicates surface flow features like wave patterns and velocities.
01:22Language Development
Children master language quickly and with relative ease, supported by both biological predisposition and reinforcement. B. F. Skinner (1957) proposed that language is learned through reinforcement, while Noam Chomsky (1965) argued that language acquisition mechanisms are biologically determined.
The critical period for language acquisition suggests that the ability to acquire language is at its peak early in life. As people age, this proficiency decreases. Language development begins very...
The critical period for language acquisition suggests that the ability to acquire language is at its peak early in life. As people age, this proficiency decreases. Language development begins very...