Today, artificial intelligence is so much more than an experiment for financial services institutions: it is a true competitive differentiator that can help banks deliver hyper-personalised customer experience, improve decision-making and boost operational efficiency.
On the final day of Money 20/20 USA, a panel of experts concludes sessions with a discussion on infusing AI across the value chain while preparing for upcoming evolutionary neural network technologies and keeping pace with new entrants.
Daragh Morrissey, director artificial intelligence, Microsoft WW Financial Services, Microsoft highlights that in the world in which we have seen the advent of self-driving cars, while the technology is developing quickly, “the regulations and the legal implications are not clear yet.”
This is linked to the question of trust and Samik Chandarana, head data & analytics, applied AI & machine learning, J.P. Morgan Corporate & Investment Bank, J.P. Morgan suggests that this will be “one of the big challenges as we move through this AI journey within financial services.”
Babak Hodjat, VP of evolutionary AI at Cognizant, uses an anecdote to explain how we are “not quite there” with voice assistants and we are not sure of where exactly to apply AI.
After Morrissey explains that it is not about one AI tool, it is about how these are combined with one another, Chandarana adds that when inside a company, “I don’t really care about this stuff and it’s really about the problem statement you’re trying to solve.
“Whether it’s about deep learning or a decision tree, it doesn’t matter. It’s about getting the right outcome and when that outcome is found, it lends itself very well to all the production processes across the field more than anything else.”
Using reconciliation as a use case, he goes on to explain that the next step is ensuring that AI is robust enough to allow the machine model to take on the task, rather than humans continuing to complete the job, very dissimilar to trading.
Morrissey adds that AI can help with the telemetry of data in regard to customer experience, advanced voice assistants and also help humans within the bank, such as the relationship advisors like mortgage advisors.
“Financial crime is one of the hottest use cases for AI right now in this industry and it’s one place where you can play safely and AI can augment the existing financial planning solutions that are there, so reducing false positives for example,” Morrissey says.
Comments from the Money 20/20 community
Gil Bolotin, VP of business development in North America for Anagog tells Finextra that we are currently witnessing the next wave of AI, “which involves placing machine learning algorithms directly onto edge (i.e., hardware) devices, such as mobile phones.
“This new branch of technology has opened the door to delivering a hyper-personalised customer experience. Today, banks struggle to understand future customer needs and preferences because they primarily rely on purchase history and online data, which has been proven to be doing a poor job predicting a poor predictor of customer needs,” offering a more holistic view of the customer.
This ties into thoughts from Michael Goodman, VP, data & analytics, and technology advisory practice lead, NTT DATA, who says that “it is critical for financial services companies to drive toward becoming more data-driven organisations, while also accelerating efforts to integrate AI and machine learning across the value chain.
“This will require organizations to develop a new approach to analytics, as use of advanced analytics historically has been concentrated in just a few organisations. In addition, they must enhance their data management capabilities to leverage the increasing variety of sources and types of data.
“Without a focus on data, they risk a breakdown where there is a lack of confidence in business-critical models, which are driven by a lack of confidence in the data used to train them. Enhancing model risk management capabilities will also help avoid regulatory concerns, as regulators are still coming to terms with AI and ML.”
Dondi Black, senior manager, payment strategies at FIS states that “AI is no longer optional or a ‘best practice’. It is critical to financial service providers who want to remain relevant and competitive in the marketplace of the future. AI enables financial service providers to make their data more actionable.”
However, while Eytan Bensoussan, co-founder and CEO of NorthOne agrees, he surmises that “where many banks stop short is one of the most important applications that AI can have for our generation: empowering people and businesses to make smarter and safer financial decisions based on their individual context and history.
“Additionally, there's a clear application of AI to understand features and use cases that could have an outsized impact on the financial health of account holders by eliminating time-consuming and burdensome manual processes. Without this, we'll only see cycles of debt and impoverishment continue in an age we have the technological capabilities to make lives much, much better,” he says.
Similarly, Mike de Vere, chief operating officer at Zest AI mentions that: “Lenders have to know whether their model is unfairly impacting protected classes (race, gender, age) or making decisions that will hurt their business. It’s essential to be able to interpret both the inputs and outputs of an AI/ML model to ensure it’s not perpetuating bias or going off the rails.
He continues: “Banks and lenders have also had to rely for years on generic scorecards that aren’t well-tuned to their specific markets and don’t get updated very often. Smaller banks and credit unions, especially, have been beholden to the pace of innovation set by their core technology providers,” which is where automated machine learning comes in.