The Centre for Data, Ethics and Innovation (CDEI) has published its AI Barometer – an analysis of the most pressing opportunities, risks and governance challenges associated with artificial intelligence (AI) and data use within the UK. The Barometer, according to CDEI Chair Roger Taylor, provides a “system-wide view of how AI and data are being used.” It explores where the opportunities lie, identifies barriers to adoptions and offers productive solutions and suggestions to improve the implementation of AI across all facets of society.
The report focusses on five key sectors: criminal justice, health and social care, digital and social media, energy and utilities and, of course, financial services. The AI Barometer devotes 20 of its 152 pages to the uptake of AI within financial services. While a large proportion of the chapter focusses on applications and AI that affects customers and citizens, within it lies some key messages for financial service providers and regulators.
We take a look at the key take aways from the report.
AI within the realm of compliance
The AI Barometer acknowledges the potential for AI within financial services, especially within compliance. Across the Industry, AI is being utilised not only by the financial institutions (FIs) themselves, but by the regulators who supervise and monitor them. The Barometer notes, “financial firms can employ machine learning techniques to comply more accurately and efficiently with regulatory requirements. Similarly, these approaches can be used as part of supervisory technology to support regulators in monitoring compliance.”
AI is a transformational technology across the realm of compliance – from regulatory reporting, to regulatory change management, to supervision and operational effectiveness. As the CDEI acknowledges, AI allows financial institutions to “allocate resources more efficiently and reduce operating costs, potentially leading to lower prices for customers”.
However, the potential of AI within FS reaches far beyond simple efficiency and cost management. While AI absolutely can allow firms to carry out usually labour-intensive processes in a fast, more cost-efficient way – its benefits are potentially limitless. As the CDEI points out, “our expert panel believed that this technology, deployed responsibly, could reshape markets for the better”
AI for cyber – a double-edged sword
We turn then to the specific areas in which the Barometer highlights the potential of AI; unsurprisingly, cyber features high on the list. AI within cyber security is a double-edged sword insofar as it presents risks, as well as resolves them. As the report notes, “increased use of data and AI within financial services and markets increases the risk and impact of cyberattacks, which may cause changes in system functionality, loss of system availability, or data breaches.”
The use of AI can undoubtedly increase opportunities for cyber-attacks and new systems might be vulnerable to new types of threats. However, the use of AI can also increase an FI’s ability to monitor, detect and prevent such threats. Moreover, the use of AI within cyber security “leads to faster and more accurate detection of cyber threats, and improved capability to counter those threats”.
Another issue presented by the AI Barometer is that “attackers tend to be ahead in the development and use of new technology”. While FIs, regulators and other businesses are often limited in innovating within the bounds of the law – cyber criminals are able to operate without these boundaries. This can mean that FIs are “on the back foot, responding to new threats rather than anticipating them”. Firms and regulators that are still relying on outdated or manual systems stand little chance of preventing or managing cyber threats when compared to those that have invested in up-to-date, AI driven solutions.
AI for AML – still in its early stages
The AI Barometer places a keen focus on the potential for AI within the sphere of anti- money laundering (AML).While AI-based AML solutions are in use across the industry, the CDEI notes that such technology is still in its early stages and commonly focusses on the assessment of individual transactions and behaviour. The Barometer suggests that there is scope for the technology to develop further so that AI is able to identify wider patterns across multiple transactions and accounts. It encourages continued evolution and exploration within this field.
The regulators lack support and resources
Over the past few years, financial regulators have changed their approach to AI and transformational technology; where once they were indifferent, they are now actively encouraging firms to innovate and embrace technology. Even more interestingly, regulators are beginning to interweave AI into their own supervisory efforts.
However, the Barometer notes that while regulators have come a long way they “may need more resources to manage the presented by AI and novel data use. ‘Where do such resources come from?’ you ask…according to the CDEI, they come from the government.
The AI Barometer notes that one of the overarching messages of the report is that there are a number of “harder to achieve” opportunities with enormous potential, but which are “unlikely to be realised without concerted government support and clear national policy”. Moreover, the report identifies “data governance and the regulatory structures” currently in place as a “common barrier” to the successful adoption and implementation of AI. Perhaps one of the pivotal messages from the report is that:
“As technology advances, so must the remit of regulators. However, their capacity and resources do not always grow in tandem with their extra responsibilities. In financial services, regulators are having to respond to a variety of new issues, including cryptocurrencies, cybersecurity threats, and a shift towards cloud-based services – as well as increased AI and data use.”
If the regulator is going to use, support the use of, and effectively regulate the use of artificial intelligence and other smart technologies, it will need the resources, expertise and technical understanding to do so. CDEI’s Barometer suggests that, at this time, the regulator is not yet fully equipped to manage the challenges and harness the full potential of AI.
AI is nothing without operational effectiveness and clarity
The AI Barometer is unquestionably in favour of technological development but doesn’t shy away from the challenges that artificial intelligence presents (see above). The key message from the CDEI is ultimately that AI should be embraced but will remain superfluous if it is not operationally effective.
“A big challenge for financial institutions lies in their ability to show regulators that they are using technology that is effective, not simply using technology because it reduces costs.”
FI’s must not only implement new technology but must be able to successfully demonstrate that such technology is operationally effective. The regulator will want to see results and improvements, not simply greater profit margins.