The FCA’s Consumer Duty expectations herald a well-embedded regulation within financial services compliance. Not only must firms now evidence customer understanding and customer support but the FCA’s intelligence-led approach will also highlight non-compliant behaviours, raising the chances of unwanted attention for firms that fail to adapt.
While it is no surprise that AI is being widely promoted as the solution to meet the higher regulatory standards, generative AI (GenAI) is not suited to the complex, judgement-based needs of compliance management and is adding rather than reducing risk. However, combining industry-specific predictive AI with the efficiency of GenAI, can be transformational. Accelerating reviews, flagging vulnerabilities and building robust audit trails ensure processes can withstand regulatory scrutiny. Furthermore, this insight also provides compliance teams with the evidence to prioritise areas of education, process change and product awareness throughout the business.
As Joe Norburn, CEO at TCC Group explains, by deploying AI that augments compliance teams, rather than working against them, firms can meet regulatory expectations, safeguard their reputation and provide a foundation for incremental business value.
Show me, don’t tell me
Across financial services, firms are under growing pressure to demonstrate the delivery of fair value to meet the FCA’s Consumer Duty expectations. The ability to evidence compliance is now as important as compliance itself, with regulators increasingly asking firms to prove that customer outcomes are being prioritised. For many, the evolution from telling the FCA about processes and training towards demonstrating customer outcomes is proving difficult. Not only is the culture completely different but processes remain manual, resource-heavy and slow to scale: to deliver the required evidence of consumer support and understanding will require firms to actively review all customer interactions.
While it is obvious that traditional compliance models are reliant upon manual review of a small percentage of customer calls that cannot be expanded to this scale, the temptation to adopt widely available GenAI tools is also proving problematic. While GenAI can do a great job in summarising conversations, it is flawed. From hallucinations to inconsistencies, running GenAI across the same customer conversation ten times will likely produce seven or eight different answers.
Similarly, the use of the call centre GenAI tools – which is essentially a sophisticated keyword search – is also unsound, not least in the possible creation of multiple false positives. Managing the needs of vulnerable customers is one of the key areas of FCA focus, yet the simplistic assumption that any conversation including the word ‘cancer’ should automatically label a customer vulnerable is incorrect. The reference could be historic, relate to a friend or family member, or even refer to their star sign.
Predictive AI imperative: Closing the compliance gap
For compliance teams, inconsistency and errors create more problems, undermine confidence and accelerate corporate risk. The lack of an evidential trail and the inability to identify the location of the underlying information further compromise the compliant state. AI does, however, have enormous potential to support compliance teams. The opportunity for the industry lies in combining compliance expertise with technology designed specifically for financial services.
Purpose-built AI – combining the precision and accuracy of predictive AI with the efficiency of generative AI – allows firms to confirm they are delivering fair value to customers and evidence compliance. Good predictive AI models can very quickly assess customer conversations, ensuring each one includes key elements such as customer identification and customer understanding.
Predictive AI tools can assess product specific requirements, such as affordability evaluations for collections and recoveries. Critically, unlike pure GenAI models, there is no hallucination. Any inconsistencies, such as a customer citing different ages during the call, are highlighted for further human assessment. And, with a full audit trail, the exact locations of the inconsistencies or areas for investigation are also provided, streamlining the follow-up review process, supporting the effective human judgement that remains central to good compliance.
Driving value
The use of AI tools developed specifically to manage the financial sector compliance not only enables firms to meet the FCA’s requirements but also transforms the role of the compliance team. Whilst compliance has always had an indication of problems within the business, the limitation of reviewing just a small margin of customer calls has restricted the reach of this knowledge. With business leaders pushing back on such a small subset of calls, it has been difficult to achieve a culture of continual improvement.
In-depth evidence collected and analysed across all customer calls provides the vital insight into operational performance required to improve firm-wide compliance. An AI dashboard quickly reveals hot spots and areas of concern, from both a process as well as individual and team perspective. Training can be prioritised for an individual with a low compliance score. Improvements can be targeted towards a specific element of a process that is not working or a high-risk product that is not being adequately handled.
The ability to quickly and accurately review every single customer interaction is fundamentally changing the role of compliance within firms. With this insight from purpose-built AI tools, firms can build a culture of continual improvement, not only identifying and fixing compliance issues but also driving tangible business value. Are advisers remembering to cross-sell key products? What are the additional factors influencing customer retention?
Conclusion
The regulatory landscape is changing, yet many firms remain mired within traditional processes and dated mindsets. It is no longer realistic to assume the FCA will focus only on the big market players or bet on a low probability of a review. The regulator is now 100% outcomes-focused and, as such, is leveraging technology to be far smarter about targeting compliance. Firms are expected to measure and evidence customer outcomes. They are tasked with identifying and remedying problems. And the FCA is adopting a data-led approach to identify good and bad players and prioritise its activity.
The FCA is clear, there are no excuses for non-compliance. But for financial firms, there is also no excuse for failing to leverage compliance insight to gain understanding, prioritise training and review product and process strategies. Predictive AI toolsets are doing far more than reducing the risk of non-compliance; they are also providing a fast track to identifying opportunities to deliver additional corporate – and customer – value.
