Banks would benefit hugely from the data gathering and management capabilities of artificial intelligence (AI) as they gird for new ESG regulations and obligations, according to business IT consulting group CGI.
While AI also could help banks in their reporting and disclosure processes, the challenge would be in ensuring the technologies are trained on accurate data, the acquisition of which is likely to remain difficult for some time, CGI vice president and global industry lead for banking, Andy Schmidt, told ESG Insight.
In his latest exclusive comments for ESG Insight on the sustainability challenges facing non-retail lenders, Schmidt said that AI would be suited to establishing the ESG performance of loan books and bond issuers. This would be of particular help in sourcing Scope 3 data from supply chains, which all financial institutions and companies are being required to disclose.“AI can help to scour, understand and document where these players are on their compliance journey and on their ESG journey,” Schmidt said. “But data is is becoming increasingly the centre of a great many conversations that we’re having, not just because of ESG reporting, but also because of the the the outsize focus on AI, specifically generative AI.”
Good Data, Bad Data
Schmidt cited Microsoft chief executive Satya Nadella, who said recently that there is no AI without data and that if you can’t trust your data inputs then you can’t trust your AI outputs.
Last year, Schmidt listed the absence of good quality data among factors having an impact on banks’ abilities to meet their sustainability obligations. The data piece for banks is intrinsically tied into other pressures they face. These include increasing regulatory compliance, a still-tangled reporting standards space and growing investor demands to improve the carbon footprint of the companies to whom the banks lend or whose bond sales they underwrite.
In Hand
While Schmidt said that banks are fully tackling those challenges, the pace of technological change is making things harder for them. All the while, they are being reminded of the time pressure they are under from the ever-greater asset risks they face in the form of threats from extreme weather incidents and social divisions.
“We recently saw some major storms, some major tidal impacts, some major rogue waves hitting towns, hitting even military installations that weren’t previously subjected to weather of this type,” he said. “So, when you’re looking at it from a financial services standpoint, whether it’s banking, whether it’s insurance, we’re gonna see more climate related losses.”
Greater Expectations
One of the consequences of these pressures is that banks, like other institutions and asset managers, are being expected to do more with their data to drive ESG outcomes. Banks, however, are a little later to the ESG ecosystem than other institutions because, argues Schmidt, banks can only urge their clients towards a more sustainable future because; they cannot compel.
Lenders are getting to grips with the data management policies that the new obligations require. But he said the volume of data they have acquired over the decades and the way in which their data management processes have developed pose their own challenges.
“Banks collect the data but being able to reach it is something different,” he said, explaining that lenders may not have been paying full attention to data quality in the past, in part, because they have been limited in their ability to control the quality of the data they receive from clients and reporting agencies.
Past Mistakes
Likening past data management and storage practices to an “attic, rather than filing cabinet”, Schmidt said that banks may have difficulty in matching, mapping and mastering data.
“Financial statements are not always saved in the same place in bank databases, for instance,” he said. “And so, while it looks good on paper to say that you should be able to pull all these financial statements, if naming conventions, document types, reporting approaches are different, then reconciling that data becomes an enormous challenge and an enormous challenge that automation can’t quite solve yet.”
This is where Schmidt sees great opportunity for AI applications. But even that runs up against a question that has become more pertinent in the past year. Economic uncertainty and weakness have added financial pressures to banks and they are now asking what those data expectations will cost.
“It’s great to have the conversation about reporting and where the data comes from and it’s a conversation that we still need to have but practical element also needs to be considered,” Schmidt said.
Subscribe to our newsletter
