Vision to reality: Harnessing AI in banking at scale and safely

It’s increasingly clear that AI – including Generative AI (GenAI) – will play a major role in strengthening customer experience, efficiency and productivity, and this will likely be followed by disruptive impact on proposition development and risk and capital management. It’s the banks that can scale AI safely across the enterprise that will win.

It’s the potential for AI to deliver personalisation at scale that could have the greatest transformational impact. Our Vision for banking report showed that 73 percent of UK consumers feel banks’ products and services are insufficiently tailored to their financial situation. AI not only has the scope to personalise customer interactions, but also to craft genuinely tailored products for every individual.

Technology firms may have the upper hand in AI development, but incumbent banks have advantages too. Most obviously, our Vision for Banking research showed that the banks enjoy a higher level of consumer trust than technology providers. They have a growing opportunity to become Intelligent Banks where the use of AI is scaled across the organisation and underpinned by robust controls, security, and ethics.

We’re currently seeing many banks experiment with targeted applications of the new GenAI technology. A typical approach is to mobilise AI pods to deliver incremental gains across a range of focused use cases, but some firms are also looking to take AI to the next level by reimagining and transforming entire end-to-end processes such as customer onboarding, claims and other customer servicing.

This level of ambition is welcome – banks need to move fast as some of their core competitors are already raising the bar and they are also in danger of being surpassed by rapid adoption among their digital native competitors. However, the possibilities of AI are changing so quickly that even AI specialists struggle to predict its future capabilities.

So how can banks prepare for an AI-enabled future and position themselves to harness AI at scale safely in ways that will add value for customers and build strategic resiliency? Banks can get onto the front foot by preparing themselves to deliver genuine whole-enterprise transformation. Here are seven key enablers to scaling the adoption of AI.

1. Prioritisation – identifying and prioritising investment based on total cost of ownership

With numerous use cases being put forward, rigorous prioritisation based on cost-value balance is essential. This should take account of both technology and business costs, potential risks, and ongoing control requirements. A value tree approach can help with modelling and prioritisation.

2. Governance – setting up risk management, oversight, and compliance guardrails

Effective governance, controls and accountabilities are vital given AI’s potential impact in areas such as Consumer Duty, fairness, intellectual property, and explainability. Governance frameworks need to extend to third parties building AI into their offerings.

3. Path to production – developing a permanent beta mindset in the business

Getting past the proof-of-concept stage requires business ownership and appreciation of AI capabilities and a continuous process of testing and implementation that maximises the value of AI use cases while managing the costs and risks associated with new models. A defined path to production playbook is key to overcoming hurdles arising from shortcomings in awareness, culture, skills, governance, data, and processes.

4. Partnering – harnessing third-party capabilities to evolve an AI ecosystem

Banks can’t develop AI on their own – key decisions around collaboration are required. Strategic partnerships with BigTechs and FinTechs as well as opportunities to obtain more data through partnerships are key to maximising the value from AI.

5. Execution – building sustainable execution muscle

Incumbents need to establish an AI Studio or centre of excellence that stays up-to-date with latest developments, sets direction, captures lessons learnt, and guides teams across the enterprise, honing the skills and capabilities and building momentum required to scale the use of AI.

6. Culture – aligning the organisation around a shared vision for AI

Realising the benefits of AI for all stakeholders depends on enabling the whole bank to use it and value it. Communicating a clear vision; defining a framework of policies, processes and incentives; and educating boards, leaders, and the workforce are key to enablement and cultural change.

7. Data – establishing high quality records and data foundations

Obtaining and integrating extensive data, both structured and unstructured documentation, and ensuring that AI is working with high quality data is vital to optimising models, mitigating biases, adhering to ethical guidelines, and ensuring that AI creates value for all stakeholders. Appropriate pre-processing, labelling, security, and compliance are crucial to the successful and ethical development and deployment of AI.

The overall goal for incumbents should be to harness AI’s amazing potential as a force for good that benefits customers, employees, shareholders, and society. Using these seven enablers as a framework, banks have an opportunity to balance the power of AI’s revolutionary capabilities with a consistent, transparent approach that builds trust.