PwC and Microsoft: Banking Horizons event series

GenAI is not just changing how work is done, it’s enabling totally new capabilities that will create new revenues. Think, for example, about a financial advisory tool underpinned by an LLM built on vast amounts of products and market information. It could open mass-market access to the kind of bespoke advice that’s today the preserve of only a minority of customers. This would create a huge boost to financial inclusiveness.

The technology to deliver this is already available. The regulatory and compliance implications remain unresolved, however. And there are other considerations to take into account in order to align with what a bank’s customers need and with its overall brand values and strategy. Resolving these mean that exposing customers directly to generative AI capabilities is probably some way down the line. But it is coming. Banks must become adept at orchestrating multiple models – large and small, proprietary and open source – to create products that delight customers, deliver value and meet compliance requirements. That is essential to move beyond proofs of concept and into production at scale.

The scale of the opportunities created by AI for banks comes with some equally weighty risks. To address them, a responsible AI framework is a non-negotiable next step. A joint report by PwC and the Association of Financial Markets in Europe reveals that while more than 60% of firms surveyed don’t yet have a framework in place, 80% have it on their roadmap for the next 12 months. What should such a framework address? Key risks of GenAI include model explanation and transparency, data privacy and confidentiality, and potential model bias. And with multiple regulatory approaches emerging around the world, it’s essential to pursue a principles-led rather than rules-based approach to respond and adapt to what is still a fast-evolving global regulatory environment.

AI will also become an increasingly powerful way to manage risks. For example, know your customer (KYC) and anti-money laundering activities will be significantly enhanced by AI’s power to proactively spot potential triggers from vast amounts of transaction data. But it’s also essential to have the ‘human in the loop’ to understand, verify and implement outputs from GenAI.

As banks develop their AI centres of excellence, responsible AI and ethics must be at the foundation.

The scale of the climate crisis is daunting. It will only be solved through collaboration. But while the technology, finance, and science required are all – theoretically – in place, the challenge is moving fast enough to activate them effectively. With greater urgency needed, developing the regulation that will galvanise investment is crucial. The US Inflation Reduction Act has raised the bar here. Other countries, including the UK, must follow suit, quickly.

Away from the regulation imperative, efforts to make the transformative changes required face two big barriers: data and skills. There may be as many as 250,000 new green jobs in the UK alone, but the skills base needs to be considerably upgraded to turn these roles into reality. That’s a task that must be taken on by governments and businesses alike. Banks and financial services businesses also need to cultivate the talent pipeline that will enable them to support their customers through the energy transition and on the road to net zero. That includes specific green skills, but also the circular thinking, change management and influencer capabilities that can help others to accelerate their own progress.

There’s capital waiting to back the businesses that can move the needle on solving the climate challenge, but investors need accurate and comprehensive data. Initiatives to improve the data available are progressing rapidly, with banks, tech companies and other data providers collaborating to create new solutions. But there also has to be a realisation that information is never going to be flawless. Investors will need to learn to work with it. Banks can prepare by ensuring that they have a solid data foundation in place, with the right data strategy, technology platform and robust data sharing embedded across the organisation.  

There are a vast, ever-expanding number of opportunities to embed financial services into consumer and business offerings for consumers and businesses. But to fully capitalise, banks and financial services businesses must decide their role in the embedded finance ecosystem – whether as customer-facing front-end orchestrators, or back-end participants providing capabilities and infrastructure. Either way, trust and transparency with customers and partners will be vital to success.

For any bank looking to tap into the potential of embedded finance, there are four key steps to take. First, carefully select and build the right business model – either as front-end orchestrator or back-end participant. Second, develop or access strong capabilities around getting insights from data to provide seamless, personalised experiences. Third, forge culturally-aligned, trusted relationships with ecosystem partners. And fourth, stay alert to the implications of the evolving regulation of embedded finance.

All four of these steps are important and interrelated. To take them successfully, banks need to think holistically, strategically and systemically.

Banks must rethink the status quo for operational and cyber resilience. A constant race between cyber adversaries and their targets means cyber and physical resilience need to become increasingly related.

Secure migration to the cloud can be a simple step to help organisations protect themselves more effectively – provided they transform both applications and infrastructure, and are prepared to drop legacy components that often create vulnerabilities.

As well as moving decisively to harness the power of cloud, banks need to address the human perspective. This means ensuring that their cybersecurity processes reflect how non-tech-savvy consumers think and act, rather than how technologists believe they should. Participants across the financial services ecosystem must also act to drive greater knowledge-sharing between organisations about the nature and source of attacks, so lessons can be learned and acted on.

When it comes to engaging with financial services customers, AI technology will become transformative. Customer insights will go beyond segments and demographics, delivering truly personalised, proactive recommendations and information. Chatbots will be smart enough to recognise customers’ intent through natural language (and even facial cues), using insights to offer appropriate, tailored customer support and journeys, all reflecting the brand, personality, and messaging of a particular bank.

Employees will see their roles change. AI co-pilots will be increasingly able to capture and summarise information from a wide range of interactions and activities. These tools will turn this insight into actionable summaries and recommendations for next steps, and simplify the process of progressing good ideas forward. As well as enhancing the customer experience and relationships, this will accelerate time-to-market for new products.

The rapid spread of AI tools means customers and employees increasingly expect capabilities to be baked into products and services. Banks can’t afford to hang back, because their competitors won’t.  

But how do they get started? Step one is to develop use cases they can build incrementally and evolve continuously, earning greater trust as they progress. That requires a two-pronged approach. First, explore the most value-adding applications for users internally and find ways to help them work faster and more efficiently. Second, identify and develop carefully-selected customer use cases. And build these on top of trusted data and ensure that customers’ trust and expectations of reliability and confidentiality are fully maintained.