Non-financial information and its contribution to the achievement of each Sustainable Development Goal (SDG) are assuming great relevance in the business world, in which it is not enough to be economically sustainable without also being sustainable from ethical, environmental, and social points of view. An analysis of how the financial sector contributes to the achievement of the SDGs is crucial in two ways. Firstly, due to the relevance and the magnitude of this sector itself; secondly, and more importantly in our view, because of the financial leverage of the banking sector that has a mandate to facilitate the transition of all economic sectors towards sustainability, guided by the 2030 Agenda. However, despite the expectations placed on banking entities, there is a research gap on their disclosure practices and on the SDG-related information that they report. In addition, the academic literature centered on the analysis of SDG-related disclosures through artificial intelligence is very scarce. To fill this gap, the objective of our study is, on the one hand, to analyze whether there is greater homogeneity in the disclosure of non-financial information in the Spanish banking sector following the transposition of Directive 2014/95/EU into Spanish Law. On the other hand, it is to evaluate the contribution of banking entities to the SDGs. To do so, the non-financial information reports of 12 Spanish banks are analyzed, completing a comparative evaluation and using artificial intelligence to identify mentions of each SDG and its targets. The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) was also used to rank the banking entities in accordance with their contribution to each SDG. The results reflected the plurality, in both breadth and quality, in the disclosure of non-financial information and in the contribution to the SDGs. The only point in common between all the entities that were studied was the use of the GRI disclosure framework and the identification of the priority SDGs, positioning SDGs 8, 13, and 4 in priority positions. The banks with higher bank capitalization levels occupied the top of the ranking of contributions to the SDGs. Differences were presented for all other aspects, even to the point of highlighting that some entities or independent verifiers had not offered all the information. In conclusion, greater efforts to improve the quality of non-financial reporting and further development of the common regulatory framework will be fundamental for better comparability between the reports from banking entities. Furthermore, this study shows that natural language processing can be applied to better measure companies' alignment with the SDGs based on the text of their non-financial reports.
