Artificial Intelligence (AI) is likely to provide opportunities to deliver smarter solutions for end users and transform the banks processing capabilities. Last week, UBS invited leaders and experts from the financial industry, fintechs, academia and regulators come together to discuss the future of Intelligent Automation and start to build a common understanding of what a successful approach would look like. I had the pleasure of attending this event and we explored the most relevant issues covering where, when and how this will happen.
For example, what do clients expect from a ‘intelligent’ bank and how can AI help us meet these expectations? Where in AI solutions does potential liability reside, how do you identify and mitigate the risk of incidents? Which business areas in Operations are the most relevant candidates for intelligent automation and can we imagine an “operation-less” bank? The answers to these and other crucial questions will define how we, and the industry, proceeds with this promising technology.
The AI-enabled client experience: Augmenting front office applications of financial services firms with AI from a client perspective. We discussed evolving client preferences, what constitutes a ‘good’ intelligent agent and how AI will impact our business. The group discussion looked at some key aspects that needed to be in place for AI to play a meaningful role in the client experience. For user adoption it is important that any AI interaction understands context and actually provides convenience to the end user. Secondly trust is crucial, and high level of transparency would strengthen trust in AI enabled solutions. AI should enhance our own cognitive abilities in a way that enables us to make better decisions, and in order to overcome the any reluctance in speaking to a machine the group proposed to give AI agents a persona.
Human vs. AI workforces in the bank of the future: Intelligent Automation with respect to people and ethics. We sounded the drivers for automation and attempted to set its ethical boundaries, investigate the optimal human / co-bot collaboration, control risks and manage the new workforce. This is a topic that inevitably takes on a more philosophical angle, and the focus should be to define the role of technology vs. The role of people. The participants all agreed that there will still be use of people in banking, but there are some tasks where machines will outperform us. These are specifically the tasks taht involve vast amounts of data and the ability to look for patterns. On the other hand, humans will still be needed for creative problem solving, imagination and human interaction. Lastly it is of out most importance to have a solid strategy in place for human displacement. This should include a vie won the role of government and regulators as well as defined accountability.
Robust and compliant AI and smart data handling: Robustness and accuracy of intelligent agents. We debated sensible requirements from regulator and bank perspective and how to identify and manage liability. We explored the fundamental role of data and how to smartly handle its ever-growing volumes. This is an area where there is no ‘one size fits all’, but common characteristics, regardless of application area should be consistency and preparedness for handling of the unexpected. From a regulatory perspective liability should be with both the provider of data as well as the algorithm and banks must be prepared to take some risk in order to provide customer value. The mentality should be that banks should go ahead and utilize the technology rather than waiting for the regulators. In order to scale banks should also take into consideration that the machine learning is costly both in terms of computing power and talent. The sheer amount of cooling needed to run the calculations favor the use machine learning libraries provided by technology giants like Google and Facebook. Banks also need to acknowledge that the necessary talent does not necessarily work for a banks, and may never want to work for a bank. In order to solve this banks should look for startups and the academic community for collaboration on machine learning.
The Automation Roadmap in Financial Services: Banks’ journey to Intelligent Automation. We prioritized tasks on the automation agenda, and looked at how operations will transform over time. From an approach where automation is used on a point solution perspective to platform-wide automation. For banks looking for where to start, they should look for areas that are data rich and decision poor as well as areas that are supportive and providing convenience before moving on to transformational areas with high uncertainty and complex decision-making processes.
No matter which part of the bank you wish to improve with the use of intelligent machines, the event feedback was unanimously: Start now!
The end result will be presented as a white paper at UBS Future of Finance Forum in January next year. Thanks to my friends at UBS for inviting me to this closed event with some of Europe’s most brilliant minds in the broad field of AI.