Agentic AI in Financial Services

This article was authored by the AI Sub-Working Group, based on insights from the Speaker Series hosted by the Emerging Technology and Innovation Working Group. More expert-led sessions are planned as part of the series.

Agentic AI in Financial Services

Insights from the Irish Funds Speaker Series

1. Introduction and Context

As part of the Irish Funds Speaker Series, industry leaders and practitioners recently convened to examine the implications of Agentic AI for financial services. The discussion highlighted a clear consensus: the industry is entering a new phase of Artificial Intelligence, where systems are no longer limited to generating outputs but are increasingly capable of executing complex, goal-driven workflows and making decisions.

This discussion from our speakers note that the evolution comes at a time of accelerated technological change, where innovation cycles are measured in months rather than years. While the opportunity is widely acknowledged, many organizations are still at an early stage of translating this potential into scalable, enterprise-wide value. The resulting tension between ambition and execution is now a defining feature of the current landscape.

2. From Generative AI to Agentic Systems

Agentic AI represents a material step change from earlier forms of automation. While Generative AI (ChatGPT, Claude etc.) is trained on the various date sources. Agentic AI will go beyond, with the capability to interact with tools, orchestrate workflows, and make decisions within defined parameters.

Speakers note that in practical terms, Agentic AI enables the decomposition of complex objectives into structured tasks, the coordination of multiple actions across systems, and the iterative refinement of outputs. Importantly, these systems operate within controlled environments and are designed to augment, rather than replace, human decision-making.

For financial institutions, this distinction is critical. The shift is not toward fully autonomous systems, but toward intelligent orchestration of work, with humans retaining oversight at key decision points.

3. Redefining Operational Models

One of the most significant impacts of Agentic AI lies in its ability to transform how work is performed. Traditional automation approaches have focused on improving individual process steps, often resulting in incremental efficiency gains but limited end-to-end impact.

The discussion at our event cantered on how Agentic AI enables a fundamentally different approach. Processes can be redesigned holistically, with multiple agents operating in parallel, sharing information, and coordinating activities dynamically. This allows organizations to eliminate structural bottlenecks rather than simply shifting them within the workflow.

In areas such as client onboarding and KYC, this shift has the potential to reduce processing times significantly while improving consistency and scalability. The implication for operating models is profound: organizations must move from process optimisation to process reinvention.

4. Value Creation in a Regulated Environment

The speakers discussed that whilst cost efficiency remains an important consideration, the discussions emphasized that the value of Agentic AI extends well beyond traditional productivity metrics.

For financial services firms, the most meaningful benefits are likely to arise from improved decision quality, enhanced risk management, and better client outcomes. In regulated environments, the ability to identify and address issues earlier in the process can materially reduce downstream risk and operational exposure.

As a result, organizations should adopt a broader value framework that incorporates financial, operational, and risk-related dimensions. This is particularly relevant in the Irish funds industry, where trust, transparency, and compliance are central to long-term competitiveness.

5. Governance, Risk and Regulatory Alignment

The transition to Agentic AI introduces new governance considerations. As systems move from providing recommendations to initiating actions, the focus shifts from explainability toward controllability.

Firms must ensure that robust guardrails are embedded from the outset, including clearly defined boundaries for system behaviour, real-time monitoring capabilities, and escalation mechanisms for high-risk scenarios. At the same time, accountability structures must remain unchanged: boards and senior management retain ultimate responsibility for all decisions, including those supported or executed by AI systems.

Regulatory expectations in this area are still evolving. However, it is clear that supervisors will expect firms to demonstrate a strong understanding of how these systems operate, how risks are managed, and how staff are trained to interact with them effectively.

6. Implementation Realities and Critical Enablers

Despite the strong momentum behind Agentic AI, successful implementation remains complex. A number of recurring challenges were highlighted during the Speaker Series.

First, data quality continues to be a foundational constraint. Without reliable, well-governed data, even the most advanced AI systems will fail to deliver consistent outcomes. Second, many organizations struggle to prioritize use cases effectively, often pursuing initiatives that are technically feasible but strategically marginal.

Experience suggests that successful organizations focus on a small number of high-impact use cases aligned with core business objectives. They invest not only in technology, but equally in people, skills, and ways of working. In particular, workforce transformation and AI literacy are emerging as critical success factors, as roles evolve from task execution toward oversight and orchestration. Crucially, strong C-suite leadership is fundamental to successful AI adoption, providing clear strategic direction and ensuring alignment between AI initiatives and business priorities. Without this leadership, organisations often fail to scale AI and embed it into core processes.

7. Strategic Implications for Irish Funds Members

For members of the Irish Funds industry, the implications are both immediate and strategic. In the near term, firms should focus on building the foundational capabilities required to support Agentic AI, including data readiness, governance frameworks, and internal expertise. Strategically, this requires business and technology teams to operate in lockstep, with processes evolving in unison

At the same time, there is a need to approach adoption in a disciplined and pragmatic manner. Not every process will benefit from Agentic AI, and the risks associated with inappropriate deployment are significant. Firms should therefore prioritize use cases where the value is clear, the risks are manageable, and the outcomes can be measured effectively.

In parallel, organizations must consider the broader ecosystem implications, including vendor dependencies, intellectual property protection, and exit strategies. Ensuring flexibility in technology choices will be essential in a rapidly evolving market.

8. Outlook: From Experimentation to Embedded Capability

Looking ahead, the speakers note that the trajectory is clear. Agentic AI is expected to move rapidly from isolated pilots to embedded capability within core business processes. Over the next twelve months, it is likely to become an integral part of how work is performed across financial services organizations.

However, the pace of adoption will vary. Firms with strong data foundations, clear strategies, and effective governance frameworks will be able to scale more quickly. Others may face delays as they address foundational gaps.

For the Irish funds industry, this creates both an opportunity and a challenge. Early movers have the potential to achieve meaningful competitive advantage, while lagging organizations risk falling behind in efficiency, service quality, and innovation.

Conclusion

Agentic AI represents a fundamental shift in the operating model of financial services. Its impact will not be limited to incremental efficiency gains, but will extend to the redesign of core processes and the redefinition of roles across the organization.

To succeed in this environment, firms must adopt a balanced approach that combines strategic ambition with disciplined execution. This includes investing in data and technology, strengthening governance frameworks, and equipping their workforce with the skills required to operate in an AI-enabled environment.

As highlighted in the Irish Funds Speaker Series, the question is no longer whether Agentic AI will transform the industry, but how quickly and effectively organizations can position themselves to capture its full potential.