This question is at the center of Tabea Krauss’s, research project, which uses AI-based text analysis methods to examine how firms disclose carbon-accounting elements such as internal carbon prices, the use of carbon credits, emissions quantification approaches, and anticipated financial effects. Krauss, PhD researcher in Law and Economics, recently visited the Center for Legal Data Science (CLDS) at the University of Zurich and tells us more about how this visit helped her strengthen both the conceptual and methodological foundations of her project on carbon-accounting and carbon-pricing disclosures under the CSRD.
Analysing climate disclosures with AI
Krauss’s project titled “Decoding Non-Financial Reports - A Natural Language Processing Analysis of Carbon Accounting Reporting under the EU CSRD” uses a dataset of around 500 sustainability reports from the first wave of companies reporting under the CSRD to understand how companies assess and value carbon emissions and climate risks. Her project highlights how legal data science can contribute to current questions in sustainability regulation. As European reporting requirements become more detailed and corporate reports become longer, computational methods can help identify patterns that would be difficult to detect through manual analysis alone.
During her stay in Zurich, Tabea collaborated with Professor Tilmann Altwicker and his team at the CLDS. They focus on using data science to better understand law in practice and to contribute to evidence-based legal research and regulation. During the visit, she further developed her project by experimenting with RAG pipelines, zero-shot classification approaches using large language models, and supervised classification through LLM fine-tuning.
Fostering collaboration and methodological exchange
The visit illustrates the importance of supporting early-career researchers in building international networks and developing advanced methodological skills. Reflecting on the visit, Tabea highlighted the value of discussing her research, code, and data with an interdisciplinary team of experienced researchers: “I am truly grateful that Erasmus School of Law supported my research visit, allowing me to pursue my research interests in collaboration with such an excellent team of researchers. The time in Zurich has not only helped me develop my research and methodological toolkit further, but also created valuable opportunities for future collaboration – and the evening swimming sessions in Lake Zurich during those hot summer days were definitely another highlight.”
Professor Tilmann Altwicker, former President of the European Society for Empirical Legal Studies (ESELS) and Chair of its Computational Methods working group, also emphasises the importance of cross-institutional methodological exchange: “Computational methods such as NLP are developing very quickly, and this makes regular exchange between researchers and institutions especially valuable. Through the Computational Methods working group, we aim to foster this kind of methodological exchange. Tabea’s visit was a great example of how networks like the European Society for Empirical Legal Studies can bring together different perspectives and help researchers develop their projects further.”
The visit is a great illustration of what internationalization can achieve. Philip Hans Franses, Vice Dean of Impact and Reputation, and Ilva Putzier, coordinator of internationalization, underline the value of international research exchange: “Supporting research stays at leading international institutions allows our PhD researchers to deepen their methodological expertise, test their ideas in interdisciplinary environments, and bring new perspectives back to Erasmus School of Law.”
- PhD student
- More information
Want to know more about the Center for Legal Data Science (CLDS)? find it here.
