Why AI requires a cultural shift in the Securities Services industry?
To make effective use of artificial intelligence (AI), banks need to take a people-first approach.
In our daily lives, we use AI almost without thinking; the algorithms in our apps tell us everything from what to buy, where to eat and where to go on holiday. But the advance of AI and other related technologies such as machine learning (ML) has been slower for many businesses – especially in the securities services sector.
The slow take-up reflects a paradox about AI: while most business executives now see it as critical to their business future, there is still uncertainty about what it is and how to deploy it. Moreover, there is widespread misunderstanding about how AI works and its implications for companies and society. In particular, there are concerns that AI could destroy jobs and even eventually challenge the dominance of humans. Some of these concerns are valid, though others are closer to science fiction.
People tend to have inflated expectations of what AI can deliver, which can lead to disappointment. The reality is that AI is a collection of innovative technologies, each with different functionality and specific uses.
At one end of the spectrum is robotic process automation (RPA), which is already being adopted for repetitive processes in securities services. It has demonstrated clear cost benefits and efficiency gains.
RPA can be implemented with limited operational changes: it uses existing systems and data to automate repeatable, clearly-defined processes that are carried out today by people.
RPA can also pave the way for more sophisticated forms of AI, including ML, natural language processing (NLP) and image recognition. These differ from RPA as they use unstructured data and rather than repeating a simple task, are instead able to ‘make decisions‘ based on a probability outcome from statistical data analysis. These technologies are scalable and efficient and therefore have the potential to drive exponential growth.
Where are the opportunities?
There are myriad potential uses of AI in securities services, often deployed alongside RPA. For example, image recognition can be used as part of the know your customer (KYC) process to scan a passport image before RPA is used to process the structured personal data. Similarly, AI can be used to spot patterns and improve fraud detection. Given banks’ huge data volumes, AI could serve as an important competitive differentiator by reducing error rates to deliver efficiency gains and cost savings.
Such use cases highlight one of the potential challenges for the deployment of AI in securities services: all AI technologies require readily available digital data. While the banking sector generates huge volumes of data, much of it is often in separate legacy systems and can be difficult to access and consolidate. Banks also have significant responsibilities in terms of client confidentiality and regulatory obligations that need to be respected before data can be leveraged on an aggregate basis.
While many banks may initially be focused on AI to create cost savings and efficiency gains, the potential for AI to generate revenues by developing new products and services is enormous. Banks are starting to draw insights from their massive data stores in order to create value. For example, analysis of operational data from the transfer agency business, can be valuable to the front office of an asset manager to assess the effectiveness of sales and marketing campaigns to distribute their funds, as well as for compliance purposes.
A people-first approach
It is important to recognise that innovation is not just about new technology, but also requires cultural change.
Unlike many previous innovations, AI is a general-purpose technology that will be deployed for a wide-range of uses. One consequence of this is that people need to be at the centre of any AI initiative.
As is often said, the goal should be ‘augmented intelligence’ rather than ‘artificial intelligence’.
This approach does not limit AI’s potential benefits. For instance, NLP can be used to continually check for price quotes and respond to clients automatically, allowing a human operator to focus on client service, while still retaining oversight of the pricing process.
Traditionally the view is that technology staff within financial services need to learn more about the business to improve effectiveness. With AI, the reverse will increasingly be true: the business must get to grips with new technologies. This process still cuts both ways, however. The CIO (Chief Information Officer) role in securities services will not only be based on technical expertise but increasingly require soft skills and the ability to act as a change agent.
AI seems certain to become critical to securities services in the future. Given the scalable nature of the technology, initial incremental benefits could rapidly create monumental changes. Firms that fail to investigate its potential uses risk being left behind. But as the industry embraces AI, it must also change its mindset and ensure that business leaders learn enough about the technology in order to understand how it can help their business. Until this happens, we will continue to see a gap between the use of AI in our personal and professional lives.