As cutting-edge technologies reshape the banking landscape, UK institutions face a pivotal challenge: harnessing data effectively to drive innovation-led growth. Robust data management underpins the successful adoption of AI.
As we approach early December's Financial Times Banking Summit, the financial services industry stands at a crossroads. Artificial Intelligence (AI) promises revolutionary change, yet its adoption in financial services lags behind expectations. This paradox demands our attention. Why are we seeing a slowdown in AI implementation when the potential benefits are clearly evident?
The UK's cautious approach to AI adoption isn't without reason. Regulatory pressures, particularly around anti-money laundering (AML), customer due diligence (CDD), and counter-terrorism financing (CTF), demand significant investments. Banks are caught between the need to innovate and the imperative to avoid hefty fines.
This caution comes at a cost. While banks focus on compliance and cost reduction, they risk falling behind in the innovation race. The business impact? Missed opportunities for efficiency gains, improved customer experiences, and new revenue streams.
Consider this: while traditional banks carefully weigh their options or cautiously embrace artificial intelligence, fintech startups sprint ahead, unburdened by legacy systems and outdated processes. They're using AI to deliver a bespoke, low-touch banking experience to Gen Z through a streamlined and digitalised banking process and 24/7 bespoke service via AI bureaucrats. If established banks don't pick up the pace, they risk losing market share to these nimble competitors.
Data dilemma
Over the last decade, discussions across various industries have consistently placed data at the forefront, recognising its pivotal role in decision-making, innovation, and problem-solving. As I will argue at a panel event at the conference on modernising the core, my years at Informatica have made one thing crystal clear: the success of any AI initiative hinges on the quality and effectiveness of data management practices.
There are risks of adoption associated with AI and generative AI. Common risks include privacy, organisational reputation, intellectual property, environmental impact, compliance and cybersecurity. However, the most significant risk is inaccuracy, which pertains to both the data and the outcomes. Therefore, the quality of our data and outcomes represents the most considerable risk. We all know it’s a case of ‘garbage in, garbage out’.
Ironically, that is not where the biggest investment is being made. You can see that businesses are investing more in cybersecurity. True, we might be able to transact more securely, but without investing in our data as an accelerator for AI, there is a real danger that you will make incorrect decisions at an accelerated pace.
It's not just about having vast amounts of information; it's about ensuring that this data is trustworthy, well-governed, and ethically sourced.
Many banks require a strong, high-quality data foundation for effective AI deployment. If we use poor-quality data, we're setting ourselves up for failure. I've seen banks pour millions into AI projects, only to see them falter because they didn't have the correct data foundation. Without reliable data, even the most sophisticated AI systems will falter. And that will likely be very costly, both financially and in terms of reputation.
We have seen unfortunate examples of chatbots selling cars for $1 and swearing at customers, attorneys relying on AI hallucinations and then having to cover them up, and medical misdiagnoses from mislabeled data – all these incidents stem from AI enablement using bad and poorly governed data.
It is in this space where the real work must begin. Banks need to invest in robust data management practices now if they want to harness the true power of AI and machine learning. It's about creating a single source of truth, breaking down data silos, and ensuring data integrity across the organisation to drive valuable and transformative AI business outcomes.
Slow progress
Progress is being made, but it's slow. Banks that have invested in machine learning capabilities for trend detection and cross-selling are better positioned to adopt AI. Yet, we're not seeing the enormous investments in data capabilities that the situation demands.
The challenge is multifaceted. Banks often need help with siloed data, inconsistent data formats, and legacy systems that don't play well with modern AI technologies. Overcoming these hurdles requires financial investment and a fundamental shift in how banks think about and manage their data.
Additionally, banks have to consider the importance of responsible AI. These systems, if not carefully designed and monitored, can perpetuate discrimination and exclusion due to the biases held by their developers. This isn't just an ethical concern; it's a business risk.
Banks must prioritise diversity and inclusion efforts across their organisations to mitigate these risks. Model governance, model risk, and model risk management are not just buzzwords – they're essential components of a robust AI strategy.
Implementing responsible AI isn't just about avoiding adverse outcomes but building trust with customers, regulators, and the wider public. In an era where trust in financial institutions is more crucial than ever, a commitment to ethical AI can be a significant competitive advantage.
Regulatory tightrope
The regulatory landscape adds another layer of complexity. Banks are under immense pressure to comply with stringent regulations. Some major banks have thousands of employees dedicated to compliance-related tasks. The fines for non-compliance are staggering – we're talking hundreds of millions of pounds.
AI can help navigate this regulatory maze more efficiently, but banks must carefully manage their implementations to ensure compliance at every step. This process requires close collaboration between AI teams, compliance officers, and regulators.
Moreover, as AI becomes more prevalent in banking, we can expect new regulations specifically targeting AI use in financial services. Banks must stay ahead of these developments, proactively engaging with regulators to shape sensible, innovation-friendly policies.
Undoubtedly, AI will play a crucial role in the evolution of traditional finance, transforming various aspects of banking. Hence, a solid data foundation is business-critical.
The benefits of a fully firing AI at a bank's core are manifold; excitingly, these are within grasp. In an environment where revenue growth is challenging, AI offers powerful tools for cost reduction, from automating back-office processes to optimising risk assessment. AI systems can process loan applications in minutes and handle most customer queries without human intervention, leading to significant cost savings.
Personalisation and predictions
Elsewhere, AI-driven personalisation and predictive analytics will transform customer interactions, improving satisfaction and loyalty. By analysing customer data, AI can offer tailored product recommendations and predict customer needs, allowing banks to engage with customers proactively. Soon, we can expect AI-powered virtual assistants to provide round-the-clock support, blurring the line between AI and human customer service.
In risk management, AI's ability to process vast amounts of data will revolutionise fraud detection and credit scoring. By analysing alternative data sources, AI can make more accurate credit decisions, potentially opening lending opportunities to underserved populations. AI will also become a powerful ally in meeting regulatory requirements more efficiently and automating many compliance processes.
Collaboration with regulators and strategic partnerships with technology providers will be crucial. Banks should start with small pilot projects to demonstrate quick wins before scaling successful initiatives across the organisation.
Ultimately, the AI revolution in banking is unstoppable. The banks that will thrive are those that act now – investing in the proper infrastructure, upskilling their workforce, and embedding ethical considerations into their AI strategies.
As the financial services industry progresses to thrive in the AI era, banks must innovate ethically, inclusively, and in a human-centric fashion. And robust data management is the foundational block on which everything else has to be built. In the AI-powered future of banking, data isn't just king – it's the entire kingdom.
To find out more about our recent customer successes and what we have to offer, visit our event hub pages:
- Cognizant hub: https://banking.live.ft.com/page/3903129/cognizant
- Informatica hub: https://banking.live.ft.com/page/4022146/informatica