Artificial Intelligence, in all of its different forms, isn’t new in the financial services space. As is often the case with new technologies, banks were early adopters of AI, with early use cases being focused on chat bots, RPA, as well as fraud and risk management. However whilst, according to McKinsey, 60% of firms have at least one AI capability deployed, to date many organisations have struggled to scale deployments. Over the coming years it is predicted that will change dramatically, and we will see AI deployments become much more pervasive as firms look to become “AI first” businesses.
Before diving into AI in financial services, let’s take a step back and baseline what we mean when we say AI. At a high-level, AI as a field can be described as any technique that enables machines to solve a task in a way similar to how humans do. It could be leveraging Machine Learning (ML), which utilises algorithms to allow computers to learn from examples without needing to be explicitly programmed to make decisions; or it could be Deep Learning (DL) which uses deep artificial neural networks as models and automatically builds a hierarchy of data representations against them. Other common forms of AI can include Natural Language Processing (NLP), which is focused on the ability to extract or generate meaning and intent from text in a readable, natural form, or Computer Vision, which is focused on extracting meaning and intent from visual elements (such as documents, pictures, videos etc). If you want to understand more about the field of AI, this blog is a great place to start.
Focusing back on financial services, as I mentioned at the start of this blog, across the industry, many firms have already embedded a AI capability of some sorts. Typically these have been either RPA solutions to automate existing tasks/processes (i.e. how Bank of America have successfully leveraged robots), chat bots to offer virtual assistant functionality to customers (well publicised examples include Bank of Montreal, Mastercard, HSBC and Wells Fargo), ML techniques to detect fraud (such as Barclays Transact), or undertaking an element of risk management when it comes to underwriting.
The opportunities to further use AI to transform financial services are immense. Indeed, McKinsey estimates that AI could potentially deliver up to $1trillion of additional value each year. Potential use cases for AI (some of which are already being trialled) include:
- Offering financial advice or portfolio management (this is what Forrester call’s “autonomous finance”)
- Reimagining the mortgage application and servicing paradigm
- Preventing cyber security attacks (great blog on this by a colleague of mine)
- Better understanding customers, and providing tailored recommendations
- Utilising algorithmic trading for investment decisions
- Real time transaction monitoring for risk or regulatory compliance
It could be said that you are only held back by your imagination! If only that was the case though; whilst there are countless opportunities to use AI, there are also a number of barriers for firms to really take advantage of it across the enterprise. Data, talent and technology readiness are typically the reasons why AI initiatives either don’t succeed, or aren’t as widespread in a firm as they could be. According to a survey by the World Economic Forum, these are all seen as major obstacles by over 70% of organisations.
For any AI programme to be a success, data is key. This is where financial services firms typically suffer as their data is often silos across multiple technologies and teams, with analytical capabilities often focused on specific use cases. The need to transform to offer a data fabric that makes it possible to access and leverage the relevant data no matter where it resides (at the edge, in the data centre or on the public cloud) is critical. But this is itself is not the only barrier – the need to industrialise the use of tools and processes when it comes to accessing this data and also create advance AI models is key to enable scale.
When it comes to talent, training and upskilling are key. But this shouldn’t just be focused on the technology teams. Business teams also need to be upskilled in the art of the possible when it comes to AI, along with some of the drawbacks and other considerations. Awareness of AI should also be focused on the board and senior management, to make sure that they too are aware of the opportunities that AI can provide, but also the challenges and risks too.
Technology, in particular infrastructure, is key a foundation for success. When it comes to AI programmes, the ability to access highly scalable, resilient and performant services is critical. Similar to many other transformational requirements, the ability to automate and drive the infrastructure as code, and leverage “pools” of software defined resources need to be consistently available – both on-premises, but also in the public cloud. Consideration also needs to be given to how technology concepts, such as in-memory databases, accelerators and data streaming are made available at scale and cost effectively for the enterprise to leverage.
By tackling these barriers, firms will be able to leverage AI to help transform into truly digital businesses. They will be able to innovate at speed, launch new products quickly and deliver new propositions integrated across customer / colleague journeys, leveraging data and technology platforms to deliver the service customers (whether corporate or consumer) are expecting. More of this in the next blog in the series which focuses on how AI can be used to transform the customer experience.
Chris Ibbitson
Hewlett Packard Enterprise
Chris_Ibbitson
Chris is a Chief Technologist for HPE, focused on the Financial Services industry. Before joining HPE, Chris has worked at both a Global Systems Integrator, as well as at a Global Bank in a variety of senior architectural roles.
