Human judgement remains critical in FS operations

Advances in digital technology offer financial services firms multiple opportunities to reengineer processes and enhance customer experience through greater efficiencies. Artificial intelligence (AI) – beginning to revolutionise other industries – is the next wave of technology credited with the potential to improve performance, in areas from investment advice to regulatory compliance to information security.

The exploitation of AI requires the finance sector to re-evaluate capabilities integral to execution. An efficient, multilateral decision-making process is composed of three elements: an infrastructure to manage information and connect the participating entities; a mechanism for sharing information; and the ability to exercise judgement based upon information assembled. Automation has a role to play in the first two elements, through its ability to process data. But can AI replace human intelligence in the final element of the decision-making process?

For example, in banking operations, innovation has partially automated and standardised post-trade tasks, but we are still a long way from full straight-through processing. Humans are regularly required to fix trade fails, with staff wading into the realm of the machines before their judgement can be utilised. The underlying network is often faulty, incomplete and out of date, leaving staff unable to reach the appropriate counterparty at another institution to locate the missing information. Even if they could be identified, issues may still remain in sourcing the information needed to resolve the problem.

That’s not to say AI has no role in achieving efficiencies. Taskize has a developed solution that eases the workload by supporting accurate, timely decision-making, enabling more efficient exception resolution. Taskize provides a platform for dispute resolution between counterparties and multiple institutions. Users benefit from the network effect of having access to contact listings by job title, but our algorithms also learn from experience, rating and ranking the appropriateness of contacts for future tasks according to historic cases. Taskize ensures that staff have all the necessary data in a single, shared environment, but leaves the final leg to an experienced human.

AI and machine learning (ML) programmes have made progress, but there are reasons why the sector should be circumspect. It is one thing if an algorithm on Netflix recommends a film that leaves you cold, it would be another if an algorithm settled a high value transaction to the wrong account. Aside from the increasing regulatory focus on investment advisors’ fiduciary responsibility a key constraint is that the decision-making logic of most AI/ML programmes is hard to identify, isolate and explain.  AI/ML is typically used to flag potential issues, rather than replace experienced staff.

The scope for ‘automaton error’ will reduce in comparison with that of manual; however, the decreasing potential for error will still demand human skill. At a time when the finality of settlement is under threat of reversal due to fraud and market stability threatened by the consequences of malfunctioning trading algorithms, we still need humans override machines when necessary. Systems can’t tell when they’re being gamed but humans can deploy their judgement with increasing efficiency. Both humans and machines remain essential to the transformation of our financial services.