UAE banking on AI, and the results are showing

Intelligence will be the new fuel for driving good governance

By Shalini Verma

Published: Wed 20 Dec 2017, 9:03 PM

Last updated: Wed 20 Dec 2017, 11:04 PM

Every once in while, an event becomes a defining moment for an emerging technology.

On October 19 this year, His Highness Sheikh Mohammed bin Rashid Al Maktoum, Vice-President and Prime Minister of the UAE and Ruler of Dubai, announced the appointment of Minister of State for Artificial Intelligence (AI). Omar bin Sultan Al Olama made history as the first minister to hold such a portfolio. The UAE Artificial Intelligence Strategy 2031 means business when it states that the goal of using AI is to cut government costs by 50 per cent and boost the country's GDP by 35 per cent.

Executing on this vision means embedding AI into every aspect of governance to scale UAE's resources, and thus fulfil the country's growing aspirations. Intelligence will be the new fuel for driving good governance.

Implicit in this initiative is an expectation of public-private partnership. The UAE government cannot do this alone. The private sector has to be equally invested in AI. Let's look at the business sectors that tend to be early adopters of a new technology. Are banks in the UAE up for the new AI challenge?

Evidently, they are at different points on the AI adoption curve. Some leading banks have gone into quasi-execution mode. The key area wherein UAE banks have started to apply AI is customer service. Emirates NBD has been very proactive, with its bilingual humanoid robot called Pepper that is trained to converse in English and Arabic, and its voice-enabled 'EVA' that speaks to you, understands your need and guides you. Other banks have invested in a virtual personal assistant that engages with a customer through a chat interface. These services are part of their effort to attract the digitally inclined millennials.

Most UAE banks have a long-term commitment to AI, but currently they are dipping their toes into conversational AI, which involves natural language processing and supervised learning techniques employed on bots.

The second is the realm of automation. Bots are being injected into the banking process to automate manual rules-based repeatable tasks, a ka Robotics Process Automation. This will free up human resources to work on more sophisticated tasks. Banks can certainly cut costs by replacing contract staff with bots. Mashreq Bank has announced its target to reduce its workforce by 10 per cent within a year. It would mean that an army of bots would replace at least 400 employees. There is a clear return on investment for replacing back office staff with these bots.

Many of the banks need to first get their house in order before they can head down the AI road. Their business units developed sequentially, which resulted in siloed systems and islands of data. Their middleware infrastructure needs to be modernised so that they can use APIs and other modern integration interfaces for a more cohesive infrastructure. The current infrastructure of many banks requires a surgical intervention so that data meant to feed AI is available unhindered, thus laying the groundwork for AI.

The groundwork also involves building in-house capabilities such as data scientists and algorithm specialists who can build machine-learning models. Banks have set up innovation centres or labs for incubating disruptive ideas. Lastly, banks are actively engaging with Fintechs to license their solution or acquire them outright. But they have challenges.

AI is still an evolving technology. Its goal post keeps shifting as some of its capabilities inevitably become mainstream, prompting the industry to pursue yet another shiny, more intelligent machine. In essence, not all the AI dots can be connected today. It's therefore a path to new discoveries. Everyone is learning along the way. The banks are learning, the technology vendors are learning, the implementation partners are learning. They all need to put their heads together to look at use cases such as predicting churn by using image recognition to detect emotions of customers entering a branch.

During WWII, the greatest inventions in computing occurred when the military joined hands with universities. It is therefore important for banks to collaborate with disparate entities such as academia, government, and technology companies, and use AI to solve a specific business problem. New ideas will pop up at these industry intersections, even though AI has a vast body of research in machine learning techniques such as neural networks.

Banks need a more profound shift in their mindset to build an experimental subculture that has a resilient stomach for failures. They need to be like seafaring pirates of the early 18th century who were not afraid of the unknown and were natural risk takers. The pirates knew the general direction they were headed in, but did not have a full grasp of the risks involved. The teams must build qualities of creative risk taking, adapting, and openness to ideas from other industries, in order to rethink banking. When AI can deliver completely new findings that are outside the purview of our hypothesis, it will inimitably rethink banking; as Tesla has reinvented cars or Netflix has rethought television. It will take several pirates to pursue the hidden treasures in banking.
Shalini Verma is CEO of Pivot Technologies

More news from