30 percent is becoming something of a benchmark for the potential offered by AI. In September, Vikram Pandit told Bloomberg that developments in technology could see about 30% of banking jobs disappear in the next five years. Now MUFG is getting in on the act.
Japan’s Mitsubishi UFJ (MUFG) plans to automate 30% of its operations by 2014 by using robots and Artificial Intelligence for paperwork that currently deploys 9,500 employees to process.
Branch staff spends about half of their time processing documents, so the argument goes. Optimising the potential offered by AI means that staff will have more time to better serve the mass affluent and private banking customers.
Where MUFG and Pandit part company is the likely effect on headcount.
Pandit argues that AI offers the potential for reduced headcount, especially in the back office.
MUFG meantime argues that it has no excess headcount and there will not be a direct correlation between greater use of AI and employee numbers.
While the back office is the most regularly quoted example of the potential offered by AI, benefits extend in a number of areas such as product delivery, marketing, compliance and customer experience.
There is at least a growing consensus regarding the potential benefits offered by AI.
Reduced costs, increased revenue, greater fraud detection, an improved customer experience and greater customer engagement are the usual compelling examples given of the benefits of AI.
Recent AI banking announcements include:
- BNP Paribas: In October, BNPP said it was using AI for a new trade matching tool and said it would be compliant with MiFIDII.
- Deutsche Bank: In October, Deutsche completed internal testing of an IBM Watson AI cognitive system. This offers a scalable and personalised advisory model supporting customers and bank employees in over 120,000 annual internal and external processes.
- HSBC: In August, HSBC said that was working with IBM to develop a cognitive intelligence solution combining optical character recognition with advanced robotics to make global trade safer and more efficient for thousands of businesses.
On the writer’s travels attending and speaking at banking conferences around the world, one of the most repeated self criticisms by bankers is their failure to manage data efficiently and profitably.
They know that they are sitting on tons of customer data but monetising that data has been a challenge. The potential offered by machine learning, AI and advanced analytics is hugely exciting.
Perhaps one of the biggest AI myths that needs to be dispelled is that AI is something new. Banks have been using forms of AI for years, in particular in processing. The best presentation at any event I have attended this year remains one from Alan McIntyre, senior MD at Accenture. He summed it up perfectly when he said that next stage of AI in banking will be toward simple and smarter interfaces: drawing on machine learning that adapts to data and interactions to improve areas like fraud detection, and tapping AI-enabled tools (like centralised platforms/assistants or messaging bots) to better converse with and offer services to customers in the front-office.
“Relying on AI for some internal and external interactions will help elevate the customer experience and move staff to more judgment-based and higher value added roles.”
As RBI’s October issue went to press, the former Barclays CEO Antony Jenkins claimed in a presentation he gave in London that big banks have “no advantage over fintech start-ups when it comes to AI.”The major banks have a number of considerable advantages over the start-ups.
I respectfully disagree with Jenkins.
Within three to five years the major retail years are likely to deploy AI as their main way of engaging with customers; they have deep pockets. They will accomplish their ambitions just as well or better than start-ups such as the one fronted by Jenkins.