Lloyds Banking Group and Nationwide Building Society are set to roll out a generative AI (GenAI) tool that has been specifically designed for the UK finance sector.
The tool, known as FinLLM, was developed in the UK by Aveni alongside the two financial services companies, with both also investing money in the firm.
The large language model (LLM) is an attempt to offer GenAI designed for use in highly regulated UK finance companies.
LLMs are a type of artificial intelligence algorithm that uses deep learning techniques and massively large data sets to understand, summarise, generate and predict new content.
Banks are in the early stage of its application across customer services, operations, and compliance and security.
Following the completion of its core capabilities, FinLLM is set to move from research into real-world deployment. Both Lloyds Bank and Nationwide are now set to use the system, which is aligned with FCA guidance and the European Union (EU) AI Act. The tool will eventually be available to the wider industry.
An LLM for the financial services sector could, for example, provide customers with explanations of complex financial matters, as well as support banks’ decisions in relation to regulatory compliance and staff making complicated calculations.
Into the real world
Nationwide’s chief data officer, Sri Kanisapakkam, said the company worked with Aveni on the development. “We’re excited by the performance of FinLLM and the potential benefits it will bring both Nationwide and our customers, as we continually look to deliver better service and experience through the responsible adoption of new technologies,” he added.
Ranil Boteju, group chief data and analytics officer at Lloyds Banking Group, said Aveni’s FinLLM “will be a game changer for UK financial services”.
The bank first invested in Aveni in 2024, and co-created and tested FinLLM on real-life AI use cases.
Lexi Birch, head of Aveni Labs, said the company brought together the expertise of its natural language processing researchers and financial industry professionals “to shape a roadmap for delivering reliable generative AI to a highly regulated sector – where accuracy and trust are non-negotiable”.
When developing LLMs, banks are paying close attention to potential risks associated with data privacy, security, regulatory compliance and potential bias in AI-based decision-making.
In terms of use cases, Chris Skinner, fintech industry expert and CEO at The Finanser, said industry-specific LLMs could help with the security of both banks and their customers. “AI bots will work out what you can trust, e.g. verifying whether the connections you make are real or fake, which in finance is critical,” he said. “With romance and deepfake scams, the more protections that AI can deliver to recognise the issues the better.
“Looking longer-term, your agentic AI bot will be the biggest security for any trade or transaction, as it will be your digital twin with its own identity and the protections therein,” said Skinner. “It’s not there yet, but FinLLM is a good start.”
Separately, Lloyds Bank announced last month that it will build, deploy and scale AI systems using Google Cloud’s Vertex AI to build a machine learning and GenAI development platform, which more than 300 of its data scientists will use.
