Seeking Al with ‘Implementation Assurance’ & ‘Manageable Risk’

Is retail banking the only space for use of AI among banks? What about treasury, corporate banking, trade finance, syndication, etc? Dr Nilay Ranjan Singh, CEO at SBI DIFC, highlights the opportunities and challenge in these lesser prominent use cases of AI:

- 

Smriti Pandey: What are the low-risk use cases for AI in treasury, corporate banking, trade finance and syndication?

Dr Nilay Ranjan Singh: AI or generative AI is still more aligned towards the retail banking considering the scale and detail. Corporate solutions are there, but not to the extent required. Retail has the problem of volume of transaction and the problem of plenty of solutions providers. In treasury, use cases are mainly in trade automation, rates discovery and matching, delivery and execution. Further check and risk management is built in for anomaly detection and overall performance monitoring.

Credit underwriting is using AI for preapproved loans, low ticket large volume loans as well as buy now pay later loans by using AI for profiling. There are good use cases available for this. The same is extendable for personal loans, car loans and to some extent, home loans.

In corporate banking, loan underwriting is still driven by process and perception, based on metrices wherein AI can give some nudge and guidance, but AI does not propel decision. Syndication is still majorly done by specified set of loan documentation, relationship engagement, price discovery and engagement through one-to-one or road show, while AI is helping in peripheral arrangements. However, overall syndication is still not AI driven. Global syndication uses relationship management more that AI. Generative AI can be of some help in documentation.