AI is no longer a future consideration for the funds industry. It's already changing how asset managers, fund service providers, and regulators operate, and the pace of that change is faster than many firms expected.
John Bohan, Head of Business Development for Ireland, recently moderated a discussion on how AI is changing the funds industry, featuring Iain Carey (Menos AI), Niamh Mulholland (Matheson), and Fergus McNally (EY Ireland). As adoption of AI in investment management becomes more widespread, this conversation between industry insiders provides a timely look at what’s working well.
AI isn't one thing. It's a stack
One of the panel's clearest points was that AI should be understood as a stack of layers rather than a single tool: infrastructure, models, orchestration, and the applications people use day to day. For fund managers, that means choosing a model or a chatbot is only a small part of the decision. The more important question is what should be bought off the shelf, what should be configured, and what needs to be built in-house to reflect a firm's fund structures, risk appetite, and regulatory obligations.
This becomes more pressing as model providers increasingly bundle models, connectors, and workflow logic into ready-made systems. The panel noted that this shift has implications for vendor contracts, with long-term agreements carrying more risk in a market that continues to move quickly.
The gap between a good use case and a good outcome
A recurring theme was the difference between a successful pilot and a successful rollout. Firms often automate a single step in a process very effectively, only to find the return on investment falls short because the steps before and after it were not considered. The panel's guidance was to take an end-to-end view of a process, and to highlight data quality early, since firms that address their data challenges first are best placed to benefit from AI.
The panel also discussed the risks that come with scaling agentic AI, where agents may manage other agents and act with greater independence. Without effective human oversight, errors can pass through a process before they are caught. The discussion includes a real example of the consequences when source and citation checks fall short.
Regulation is developing alongside the technology
The legal perspective brought a grounded view of where things stand. Legislation and supervisory frameworks are developing at the same time as the technology itself, and questions of accountability – particularly who is responsible when an AI-supported decision goes wrong – continue to sit with the people overseeing these systems. The panel also discussed how regulators are exploring AI to turn years of collected data into more timely insight, which could influence how quickly regulatory frameworks are updated in future.
The harder question is what happens as AI becomes more autonomous. Will liability always remain with a human-in-the-loop, or will legal frameworks eventually change with the technology?
Why this matters now
AI in the funds industry is a present reality, not a future one. Firms are already investing significant resources, and the difference between a strong return and a costly misstep often comes down to whether AI is applied across a full process, the quality of the underlying data, and the controls put in place from the start. This discussion offers a practical framework for fund managers and companies working through that process.
Watch the webinar to hear the full discussion, including how firms are approaching AI adoption, where agentic AI creates new risks, how regulators may use AI themselves, and what the funds industry could look like in ten years' time.
