Banks struggle with CSRD as data gaps and AI loom

Now that the first reports under the Corporate Sustainability Reporting Directive (CSRD) have been submitted, financial institutions are facing the challenge of having to move from one-off compliance efforts to sustainable annual processes.

Research shows that many banks and asset managers are struggling with recurring bottlenecks, including inconsistent ESG data, heavy workloads, and legal uncertainty.

According to the research, the first CSRD cycle has exposed both procedural and content-related challenges. 

Many institutions report that they lack clear data definitions and lineage. Data collection is still heavily manual, with thousands of potential data points needing to be sourced and tracked.

“Most data flows still need to be established, potentially involving up to 1,200 data points,” said Rens Krediet, researcher at ACE + Company which carried out the study. 

“This creates a significant effort to understand what each data point means, where the data can be sourced, and how it moves from point A to B.”

The reliance on manual work has also raised concerns about data quality. Institutions fear that errors or inconsistencies could undermine both regulatory compliance and market confidence.

Looking ahead, Krediet argues that further professionalisation is essential for the second reporting cycle. Banks see the most potential in data automation, centralised intake platforms, and clearly defined responsibility structures with traceable flows. He added that generative AI could become an important enabler of this shift.

For now, many organisations are using transitional structures. Temporary teams are testing new tools and approaches before they are rolled out across the wider business. At the same time, reliance on external consultants is declining as banks shift more of the work in-house.

“We’re currently internalising work that was previously outsourced, so we can become self-sufficient in the long run,” said one research participant. Krediet said this shift would help embed CSRD into daily operations.

Another area of focus is the integration of CSRD into risk and control frameworks. Institutions are mapping CSRD requirements against existing policies to identify gaps and overlaps. 

Krediet noted that this is a demanding task, but one where technology can help. “With smart tools like Reg AI and Gen AI, this kind of work can be carried out much more efficiently and with shorter turnaround times,” he said.

However, the Omnibus Directive has slowed momentum. Many banks are waiting to see how the new rules take shape before committing resources to process improvements. 

According to Krediet, this may save money in the short term but risks delaying the learning curve that would improve efficiency in future cycles.

Another issue is internal buy-in. “We’re still trying to sell the importance of CSRD internally,” one participant admitted. Without strong leadership support, improvements remain deprioritised.

Institutions are also benchmarking their approaches against peers and joining market initiatives to identify best practices. Observing how other banks adapt is seen as a way to accelerate their own learning.

Krediet’s research highlights GenAI’s ability to accelerate and structure ESG reporting, support consistent data models, and automate validation. GenAI can also be used to link policy documents with CSRD requirements, flagging overlaps or gaps, and to monitor legislative updates such as the Omnibus.

With the first reporting cycle now complete, the challenge for financial institutions is to move beyond compliance box-ticking. 

The focus will be on embedding CSRD into governance and risk management, while using technology to reduce cost and complexity. Those that succeed could turn regulatory reporting into a strategic advantage rather than a compliance burden.