The rise of generative AI and Large Language Models (LLMs), coupled with the advent of Open Banking, has sparked a debate: Are we on the cusp of a new era in personalised banking, or are the risks too high to take the plunge?
At Red Badger, we’ve been at the forefront of designing and building complex digital products for major enterprises, including global banks. Our hands-on experience with LLMs offers a unique lens through which to examine the challenges and opportunities that lie ahead.
Understanding Generative AI
Generative AI is a subset of artificial intelligence that can autonomously generate content, be it text, images, or even code. Unlike traditional software, which operates deterministically, the output from LLMs is probabilistic. This makes them ideal for creative tasks but introduces risks when applied to regulated activities, like financial advice.
Enhancing customer journeys
Personalised financial advice
Generative AI can sift through a customer’s financial data and transaction history to offer tailored advice. However, the line between guidance and regulated financial advice is thin.
AI stock pickers and AI-enhanced robo-advisers are inevitable and the way customers interact with banks is set to evolve dramatically.
Servicing
LLMs can significantly improve customer service by interpreting customer intent effectively. Future LLMs could be integrated with Robotic Process Automation (RPA) systems to offer hyper-personalised service journeys, reducing customer frustrations and increasing automated completions. LLMs coupled to APIs will revolutionise service overheads and customer service experience.
Redefining user interfaces
Generative AI has the potential to revolutionise user interfaces in banking. From improved conversational interfaces to personalised, dynamic dashboards that can be customised in real-time to display financial data, spending insights, product options, and model scenarios in an intuitive way.
Financial workflows can be complex. LLMs’ ability to infer intent is a powerful new capability for customers with diverse accessibility needs to explain what they require.
Grabbing the bull by the horns
Be experimental
The landscape is new and ripe for experimentation. Starting small and learning as you go is crucial.
Our experience shows that customers are open to interacting with LLMs, provided there are clear paths to more traditional or human interactions.
No high stakes without guardrails
Ethical considerations around AI are paramount. While foundational models have reduced biases, the unpredictable nature of LLMs means that safety measures are essential.
A ‘blended AI’ approach, combining human and machine intelligence, is likely to be the most pragmatic solution.
Recognise the hype
The industry is rife with overblown expectations around AI. It’s crucial to set clear objectives and tackle the hard problems head-on, such as data privacy, algorithmic bias, and security risks.
Conclusion
Generative AI has the potential to be a game-changer in the financial sector, offering more personalised and compelling interactions.
However, challenges around market competition, consumer protection, and potential risks cannot be ignored.
As we navigate this new frontier, a balanced approach that experiments with the strengths of both human and machine intelligence will be key to unlocking the full potential of AI in banking.
John Godfrey is commercial director at Red Badger
