Banks are supposed to be sitting on masses of data which, if they could put it to work, would be gold dust. In fact they need more data to do a proper job, writes Brian Caplen.
When will artificial intelligence (AI) put credit officers out of work? They can probably feel safe for now. Despite all we are told about banks having masses of data and only needing the correct IT systems to harness it, a successful AI credit programme would need to tap a wide range of sources to emulate a credit officer in business banking.
Banks do of course have in-house data on companies and borrowers, but to really get a good result they would need to plug in sources such as the UK’s Companies House (or the equivalent government register in other countries), Dun & Bradstreet, Bureau van Dijk and so on.
These are data sources that credit officers use but which may not be integrated into an automated system. That means that any bank CEO who wants to cut costs by introducing the latest in AI would first need to get a fairly extensive IT project underway.
Herein lies a second problem which banks are increasingly coming up against. Their massive IT departments can be a source of resistance rather than a force for change. After all, slick modern tech puts IT jobs at risk too. All too often the IT department’s stance is about preserving the status quo and the budget rather than thinking outside the box. Increasingly, banks need to outsource and move into the cloud if they are really to make headway.
Says Imam Hoque, COO at data analytics company Quantexa: “We always hear about how banks should have a single view of the customer and utilise multiple data sources, but to do this properly using AI, their systems need to plug in and network many more data sources. To achieve this in a timely manner, chief data officers are often finding that the traditional relationships with internal IT may need a rethink.”
Banks have little choice in getting to work to tackle this issue. Fintechs are building their credit systems on this basis right from the outset. Banks either have to get with the programme or face losing the business.
