The Sustainable Finance Cluster has published a new position paper titled “AI in ESG Management – Adoption Maturity, Use Cases and Responsible Governance in Banking.” The publication demonstrates how banks can use Artificial Intelligence (AI) to process ESG data more efficiently, improve analyses, and provide more targeted support for reporting and risk processes. At the same time, the paper makes it clear that the banking sector has recognized the strategic importance of the topic, but is still in the early stages operationally. The crucial step now is to move from individual pilot projects to scalable, responsibly managed applications. The position paper was developed as part of the cluster initiative “AI in ESG Management” with experts from the fields of Sustainable Finance, Risk, Data, and AI.
Key findings of the position paper:
- AI has arrived in banks’ ESG management, but is so far predominantly used in isolated instances. General-purpose AI chatbots are widespread, whereas more integrated, document-based, or agentic applications are mostly still in development or the pilot phase.
- The greatest benefit is currently seen in efficiency, knowledge building, and quality assurance. AI can analyze ESG reports, extract data, prepare documentation, monitor external sources, and relieve specialist departments of recurring analytical tasks.
- Scaling remains challenging. Technical infrastructure, existing system landscapes, limited resources, lack of AI competencies, and limited trust in AI results are slowing down the transition from pilot projects to production applications.
- Responsible governance is becoming the critical success factor. When AI impacts processes such as credit assessment, disclosure, or ESG risk management, it requires clear responsibilities, human oversight, traceability, documentation, and a mindful approach to model risks as well as the environmental footprint of AI.
From pilot projects to scalable application:
The paper examines the current maturity level of AI use in ESG management, concrete use cases along the ESG value chain, and the requirements for responsible and sustainable AI governance.
- Banks should systematically scale proven AI use cases instead of constantly developing new isolated solutions. Reusable building blocks such as document analysis, retrieval, classification, and source-based text generation can form the foundation for multiple ESG processes.
- Use cases should not be selected solely based on what is technically quick to implement. Data availability, regulatory alignment, process integration, user acceptance, and a measurable value-add for ESG work are also decisive.
- Case studies from Helaba, Novata, and Commerzbank show that AI creates value particularly where it supports experts in analysis, screening, prioritization, and documentation, without replacing their final judgment.
- The next level of development lies in more integrated and, looking ahead, agentic applications. However, these should only be pursued where data quality, control mechanisms, responsibilities, and auditability are sufficiently developed.
Recommendations for banks:
- Banks should treat AI in ESG management as a portfolio and architecture issue: Standard solutions can accelerate initial adoption, while in-house developments make sense where they create reusable capabilities, control, and strategic differentiation.
- Technical foundations and collaboration must be strengthened. This includes shared ESG data structures, clear data responsibility, and close cooperation between ESG functions, specialist departments, IT, risk management, and future users.
- Institutions should align AI governance early on with the requirements of the EU AI Act, especially for decision-relevant applications such as credit processes. Human oversight, source referencing, documentation, and model risk management should not be added retroactively, but built into workflows from the start.
- Policymakers, supervisors, and standard-setters should support proportionate, practical regulatory frameworks while simultaneously improving the prerequisites for responsible scaling – for example, through greater clarity on high-risk applications, data access, standards for sustainability disclosures, and methods for assessing the environmental footprint of AI.
Matthias Hübner, Managing Director of the Sustainable Finance Cluster, emphasizes: “Artificial intelligence can help banks use ESG information faster, more consistently, and in a more decision-relevant way. However, the technology alone is not the decisive factor. The value is only realized when institutions systematically scale AI, clarify responsibilities, and design their applications in a way that ensures results remain traceable, controllable, and robust.”
Dr. Lukas Figge-Muschalik, Scientific Director of the project and the position paper, adds: “Our analysis shows: For ESG management, the question is no longer whether banks will use AI, but how they will manage the transition from individual pilot projects to responsible scaling. This requires reusable technical capabilities, clear governance, and an understanding that the risks and resource consumption of AI itself are part of sustainable financial practice.”
About the Sustainable Finance Cluster:
The Sustainable Finance Cluster was founded in 2018 through the merger of the Accelerating Sustainable Finance Initiative of Deutsche Börse and the Green Finance Cluster Frankfurt of the Hessian Ministry of Economic Affairs. The aim is to formulate and implement concrete approaches for the future viability of national and international financial markets. The cluster serves as a central point of contact for Sustainable Finance inquiries from the financial sector, the real economy, academia, and policymakers.
Note: The thumbnail was created using generative AI.