By Dorian Selz, CEO at Squirro
GenAI has the potential to be truly transformative for financial organisations across the world, which is why millions are being spent by the sector to realise its promise of increased productivity, enhanced operational efficiency, improved customer service quality and support for service and product innovations.
A survey carried out by BCG in July last year found that investment in GenAI was expected to grow by as much as 60% by 2027, and one in three companies were planning to spend $25 million+ on AI this year.
But here’s the thing: It’s not clear whether investments already made are coming to fruition, and if not, what needs to be done to make sure current spending delivers a tangible return. The BCG survey found that while 75% of executives ranked AI/GenAI as a top strategic priority, only a quarter of them were seeing significant value.
There are potentially many reasons for this, from setting goals too low and spreading AI usage too thinly over multiple activities, through to failing to properly train employees and implement cultural changes or put in place KPIs they can monitor. Many banks and finance companies also have concerns about the legal and regulatory risks of GenAI projects.
A recent report by the Swiss Bankers Association, encourages organisations not to think about short-term rewards, instead stressing that integrating GenAI into banking is about thoughtfully incorporating it into mid-to-long-term business strategies. They point out that one of the strategic benefits is the general technological maturity of the organisation it fosters, particularly by empowering employees to evolve with the technology. The favoured approach for banks, they say, would be to start with GenAI now, which will gradually build internal expertise and allow experimentation.
To increase the proportion of organisations that are seeing value means using a framework to help plan and execute their GenAI deployments responsibly, and the Swiss Bankers Association report outlines an effective approach for banks. It recommends that they act across multiple dimensions, ensuring that AI initiatives are aligned with business strategy, governed effectively, supported by the right mindset and culture, underpinned by robust IT infrastructure, and secured by rigorous risk controls. This will allow them to harness the power and multiple benefits of GenAI while minimising risks and ensuring compliance with regulatory and ethical standards.
The Role of RAG
Recognising the huge role that AI plays in managing and making sense of a bank’s data, the SBA report pulls out one significant answer to the question of increasing value – Retrieval Augmented Generation (RAG). This is a method used to find and pull together all of a bank’s data from many disparate systems so that it can be safely infused into GenAI applications. These can then understand the meaning of a query and connect related data points, revealing hidden insights and providing a holistic view. If banks can intelligently combine their understanding of what GenAI models can do in conjunction with their internal data resources, they will be able to create more accurate, relevant, and impactful outcomes.
Relieving the pain points
Banks need to look beyond shiny features and gain a deeper understanding of what GenAI can do to address their most difficult pain points. Examples of this include:
Breaking down data silos & information overload – Hunting through disparate systems for key data is time-consuming, but semantic search and GraphRAG (a form of RAG that leverages an enterprise’s taxonomy and knowledge graph) can deliver a holistic view of a bank’s data, leading to faster, better-informed decisions.
Navigating security and compliance worries –Concernsabout deploying GenAI with sensitive data can be allayed by using anomaly detection, proactive risk assessment, and automated compliance reporting to enhance GenAI security and ease regulatory burdens. Importantly, the BCG report found that data privacy and security, lack of control or understanding of AI decisions, and regulatory challenges and compliance were the three key AI risks that they had to navigate.
Prioritising the customer experience – All banks are aware of the importance of personalised and seamless interactions, and GenAI can deliver chatbots and tailored recommendations which will boost satisfaction and loyalty.
Tackling operational inefficiencies – Legacy tasks that are still being carried out manually can stop immediately when GenAI is implemented to automate routine processes like data entry, document processing, client meeting preparation and reports, with the benefit of liberating valuable human resources.
Bridging the talent and skill gaps – the common issue of finding experienced IT expertise can be eased by using GenAI to augment existing teams, speed up employee on-boarding and empower employees by delivering fast access to critical information.
Streamlining auditing – GenAI is able to support risk, audit and compliance management, helping to automate data gathering and report generation, making audits faster, more efficient, and less stressful.
Closing the AI impact gap and realising value will take a measured approach. There is no doubt about the benefits of GenAI from automating compliance checks and generating instant client insights through to accelerating document analysis and boosting efficiency and productivity. Secure deployment requires robust governance, model monitoring, and phased rollouts and as the Swiss Bankers Association points out, scalable strategies aligned with the broader business.
Banks should research actionable use cases and look at examples of how other organisations are managing their deployments to determine their own calculated, risk-aware approach. This will enable them to tangibly capture ROI through cost savings, faster service, and improved client experiences, and deploy and scale their GenAI initiatives successfully. After all, the only results that matter are those that can be proven.
