“Most executives are fully aware of the buzzwords like blockchain, artificial intelligence, data science, and machine learning. They say ‘Wow! Fantastic! We’ve got to move towards it.’ But they don’t actually know the use cases, how it can be implemented or the pre-requisites.” This is the view of Mo Haghighi who is head of developer ecosystems for IBM in UK and Ireland, giving his opinion of executives in the financial services industry.
Haghighi was sitting on a panel at the Rise London fintech innovation space discussing artificial intelligence and big data in financial services. Alongside him were Lee Baker, commercial director of open source machine learning deployment platform Seldon, and Justin Lyon CEO of Simudyne, a secure simulation software development kit.
"Picking out the most inexplicable behaviour is actually quite interesting.”
The panel was asked about possible use cases of AI and big data in banks. In response Haghighi said he really liked the idea of using AI for anomaly detection. “Looking at billions of transactions and trying to pick out the most inexplicable behaviour is actually quite interesting,” he said. He is also excited by the idea of virtual advisers that can analyse a user’s behaviour and make suggestions and recommendations about purchases and investment decisions.
Lyon of Simudyne said that AI and big data can make a big difference in the area of customer insights with the creation of digital twins. He said that “little virtual versions” of people can be created and used to form a virtual marketplace where reactions to new products or services can be tested. He said that financial institutions can get a sense of all sorts of plausible futures that can be taken back to the real world and actioned. “That’s really cool and the banks have the data. They can do some quite interesting experimentation safely,” said Lyon.
“AI that mitigates or moderates for regulation is going to become very, very significant.”
Baker came up with three use cases. The first would be the use of AI to manage compliance, saying: “AI that mitigates or moderates for regulation is going to become very, very significant.” The second is any case where a signal can be extracted from trading - this is going to make a massive difference and be a real leveller.
“Banks are under a great deal of pressure from hedge funds and systematic traders because they are just better at curation of data and applying models, so I'm interested to see where the trading space goes,” he said. Baker also said that banks can make a philanthropic difference by using artificial intelligence and machine learning in predicting credit risk.
However, the panel identified significant challenges to the adoption of these technologies. In Baker’s view some financial services executives might be held back by anxiety. Despite believing in the application of machine learning and artificial intelligence, they have concerns about data governance and data management. “They’re looking at a whole pipeline of things and they're stymied by the anxiety of where to start,” he said.
Decisions are being made by executives who are “absolutely” out of touch.
According to Haghighi, an impediment to the adoption of AI and big data is that in banks across the board, decisions are being made by executives who are “absolutely” out of touch. He said he has seen many interesting and innovative projects that have been killed off by executives before even reaching the prototype stage.
Baker disagreed that executives in big banks are out of touch, though he conceded that they may be distanced from the person in the street. He stated the biggest obstacle is the difference between the culture and methodology of the smaller business units wanting to adopt new technologies and the wider organisation. “Their business unit wants to do something sweet, nimble and agile, but that’s not the monolithic approach of the big financial services company.”
Baker elaborated on his view by stating that the ambitious goals and mentality of the big banks are incompatible with the iterative approaches needed to bring in disruptive technologies like AI and big data, thus creating a cultural tension.
"They don’t want to do the quick win, they want to do the big win.”
“Banks either want to go all or nothing. There is a cultural schism where they don’t want to do the quick win, they want to do the big win,” he said.
Lyon said that he has found resistance to disruptive technologies, not from top-level executives, but from heads of desk. He said that executive committees often “get it,” as do the developers because they are engineers.
However, he found that he had to have conversations with the people “who own business” to make sure they understand the business value and how the new technology will help them reach their target.
Few doubt that big data and artificial intelligence will transform financial services, as it will to many other industries. PricewaterhouseCoopers dedicated a chapter each to data and analytics, and AI and digitial labour in its report on top issue for financial services. It is just a matter of seeing which use case will be adopted mostly widely and the impact that will have on the financial institution and its customers.