The application of AI-driven digital transformation in financial markets is accelerating, yet firms are grappling with the challenge of translating innovation into tangible results. How are financial institutions successfully integrating AI into front-office operations? What are the biggest obstacles to AI adoption, and how can firms move beyond experimentation to achieve measurable returns? Should firms build their own AI models, or is partnering with third-party providers the more effective strategy? And critically, how can organisations ensure their AI investments translate into real competitive advantage rather than just incremental efficiency gains?
These were some of the key questions discussed during a panel session at the A-Team Group’s recent TradingTech Summit London, moderated by Monika Fernando, Head of Global FI Client Data Analytics & Head of FI eTrading Strategy EAP at TD Securities, and featuring Anvar Karimson, Chief Technology Officer at Kepler Cheuvreux; Sachin Anandikar, Chief Technology Officer at Pemberton Asset Management; and Barry Fitzgerald, Co-Head of Front-office Engineering at Man Group.
AI Adoption: Moving Beyond Experimentation
AI adoption in front-office functions is at different stages across firms, but the overarching theme is clear – while traditional AI techniques such as reinforcement learning and predictive analytics have long been used, generative AI (Gen AI) is redefining expectations.One panellist described their firm’s early foray into Gen AI: “About two years ago, our chairman asked what we were doing with AI. The answer was ‘nothing.’ The response was ‘do something.’ So we started exploring, and the first use case was all about text manipulation.”
This shift to Gen AI has been particularly effective in automating tasks like text summarisation, query transformation, and converting structured data into more accessible formats. In some firms, AI tools are already delivering personalised investment reports by parsing market news and regulatory statements for relevant forward-looking insights -capabilities that were nearly impossible a few years ago.
Despite these advances, firms are taking a measured approach when it comes to using AI for trading execution. While sentiment analysis and portfolio pricing have seen AI integration, a human-in-the-loop approach remains standard.
From Productivity Gains to Tangible ROI
A key challenge for AI adoption is moving beyond efficiency gains to measurable return on investment (ROI). One panellist shared a candid example: “We automated a workflow that significantly improved efficiency, but what did the team do with their free time? If they just took longer lunches, then there was no real ROI.”This illustrates a critical point – AI’s success is not just about automation but about leveraging those efficiencies to drive greater value. Firms seeing the most impact have targeted AI at high-value areas such as private debt analysis and legal automation. By streamlining credit assessments and NDA processing, they’ve increased throughput and reduced costs, demonstrating direct bottom-line benefits.
Building the Right AI Infrastructure
For firms looking to scale AI adoption, investment in foundational infrastructure is essential. A panellist framed it succinctly: “Will you ever regret having organised data or more compute resources? Probably not.”
AI’s effectiveness is directly tied to the quality of data it can access. Financial firms are increasingly focusing on data governance, cloud partnerships, and ensuring AI tools integrate seamlessly into existing systems. AI models require structured, well-documented data – much like a new hire needing clear onboarding material. Without this foundation, AI’s potential remains unrealised.
The Specialisation Debate: Build vs Buy
A point of contention in the discussion was whether financial firms should invest in training their own AI models or rely on general-purpose solutions. Some firms initially pursued highly specialised models trained on proprietary financial data. However, as one panellist noted, “We found that a large general model trained on everything often outperforms a narrow, finance-specific model.”
Yet, the tide may be shifting again. Advances in reinforcement learning and more cost-effective model training are making bespoke AI solutions more viable. As a result, firms are weighing whether to invest in their own intellectual property (IP) or partner with AI providers. “If you’re developing your own IP, invest in it. Otherwise, partner.”
Talent and Organisational Readiness
The talent landscape for AI implementation is evolving. While demand for AI expertise is high, firms are realising that the most valuable hires are not necessarily AI specialists but those who understand internal data and business context.
One example stood out: “We hired a junior HR team member – a Shakespearean actor by background. Within months, he built a chatbot answering HR queries. AI skills don’t have to come from computer scientists.”
This underscores the broader shift – AI adoption is no longer just a technical challenge but an organisational one. Firms need strong internal advocates, targeted upskilling initiatives, and clear leadership buy-in to drive successful adoption.
What’s Next? AI’s Evolution in Financial Markets
Looking ahead, AI’s role in financial markets will continue to expand, but so will regulatory scrutiny. The panel anticipated developments in sovereign AI models, tighter compliance frameworks, and more sophisticated blends of deep learning and machine learning.Despite the rapid pace of change, the consensus remains that AI is unlikely to replace human decision-making in front-office roles. Rather, its value lies in enhancing productivity, reducing friction, and enabling financial professionals to focus on higher-order tasks.
As one panellist concluded, “Expect change. A lot of it. The land grab for AI is real, and the firms that structure their data, define clear ROI, and build the right partnerships will be the ones that thrive.”
- A-Team Group AI in Capital Markets Summit London 2025 will be held on May 22. To see the full agenda, click here or register below.
Subscribe to our newsletter
