Fintech Forward Newsletter - Issue 15

Sarah Mason

13 March 2025

With AI adoption growing across the financial services industry, assessing the practical use cases and the tools that will help deliver valuable AI-powered solutions and real transformation is crucial. With so many Gen AI solutions and large language models (LLMs) to choose from today, IT decision-makers must seek to understand the subtle and stark differences between what’s on offer and determine the use cases that can benefit from each offering.  

A use case highlighted by IBM is customer risk assessment for banks when it comes to recommending financial products. Using generative AI to analyze market trends, financial indicators and credit histories, banks can deliver saliant information to analysts and accelerate decision making. Again, however, choosing the right LLM for the use case will be essential. 

We also take the opportunity to flag an extensive report from The Alan Turing Institute, authored by finance industry experts, including representatives from banks, the FCA, Accenture and analysts, about how to ensure the trustworthy adoption of LLMs.  

We also consider why one LLM may not be enough to achieve the objectives of AI projects and AI-led transformation. With different LLMs excelling at different tasks, and some having multimodal capabilities, AI strategies are growing in complexity. With this week’s newsletter, we hope to unravel some of that complexity.  

A modern approach to customer risk assessment with LLMs

The risk assessment landscape is evolving, with a growing demand for accurate evaluations. Gen AI is helping teams to target customers more effectively whilst multimodal large language model pipelines assist banks in assessing and explaining product recommendations and customer approvals. 

The Alan Turing Institute: Trustworthy adoption of LLMs in finance

Firms are adopting LLMs internally and evaluating their market potential. Auditable models are needed to reduce risks, alongside addressing AI concentration in large firms. Recommendations include sector-wide best practice analysis and exploring open-source financial AI models like FinMA and FinGPT. 

The value of adopting multiple LLMs

As businesses assess the value and use cases of LLMs, IT business decision makers should not shy away from making the case for adopting more than one. With so many to choose from, and each having different strengths, it’s essential to consider how LLMs might complement one another and accelerate success with AI 

Top Tip

Is multimodal AI the future?

Artificial Intelligence has progressed beyond handling text and the future lies in multimodal AI, which refers to models that can simultaneously process text, images, video and audio. But what makes this development so important? 

Webinar

Transforming wealth management through macroeconomic insights

Serving and satisfying wealth clients demands highly relevant and personalized services. This upcoming Finextra webinar featuring JBI Metia, explores the use of macroeconomic data and AI to create value for investors and high-net-worth individuals. 

Blog

Gen AI, agentic AI and I: The changing face of content 

Forecasts predict that by 2026, 90% of the content available on the internet will be generated by AI. Marketers who invest in teams capable of developing LLMs will stand out. These models must capture essential messaging in a distinct tone and align with clear marketing objectives. 

Report

B2B Marketing Trends Report 2025

Read Metia's report featuring seven key trends to learn how to monetize your AI strategy, harness intent data, and build greater affinity with your customers.