Artificial Intelligence (“AI”) Insights from the April 2025 GAIM Ops Conference
GAIM Ops 2025 offered an exciting look at how AI is being applied in investment operations, revealing both promising opportunities and challenges that firms must navigate.
AI Risk Takes Centre Stage
The AI Summit held during the opening Sunday of the conference started with a series of informative discussions about the implications of AI risks, good governance and responsible use. One presenter revealed concerning data indicating that AI-generated phishing emails have significantly higher success rates compared to traditional methods. Particularly noteworthy are the advancements in sophisticated voice and video deception techniques, which pose challenges to conventional security measures through the application of AI.
Several presenters emphasised the critical role of third-party risk management, particularly as organisations increasingly share sensitive internal information with external vendors for model development within complex supply chain ecosystems. The scope of risks has broadened substantially, encompassing not only data breaches but also privacy infringements, intellectual property theft, and the intentional dissemination of misinformation via AI platforms.
Multiple speakers underscored the pressing need for comprehensive Board education on these emerging threats, to ensure that leaders remain informed and prepared to navigate the swiftly changing threat landscape.
Practical Applications Delivering Results
The conference showcased compelling real-world implementations that moved beyond theoretical possibilities. One investment firm detailed their AI coding assistant deployed to over 300 developers with impressive productivity results. Another use case highlighted an information retrieval system that allows staff to access expert-level knowledge through natural language queries.
Current applications across the industry are remarkably diverse: enhancing research capabilities through superior pattern recognition, automating portions of client onboarding, simplifying due diligence questionnaires, and generating customised RFP responses. Sophisticated implementations discussed include sentiment analysis of analyst calls to identify subtle market signals and hybrid workflows that combine AI efficiency with crucial human oversight.
Regulatory Response Progressing
A March SEC roundtable on AI featured prominently in discussions, reflecting growing regulatory interest. Speakers acknowledged the persistent gap between technological advancement and regulatory frameworks, creating compliance challenges for forward-thinking firms.
This regulatory uncertainty has prompted some organisations to create dedicated AI compliance positions to develop internal governance frameworks and speakers commented on the emergence of the AI compliance officer as a distinct role with specific technical and ethical expertise.
Efficiency Gains Driving Adoption
Financial institutions are increasingly viewing AI as a powerful multiplier for human capabilities, particularly for data-intensive operations. Practical applications gaining traction include extracting structured information from unstructured documents, streamlining client onboarding processes, and creating intuitive interfaces for complex reporting systems.
Industry projections suggest billions in potential cost savings for asset managers by 2030 and speakers emphasised that the next 24 months will focus primarily on efficiency and automation. Particular attention is being given to routine tasks like drafting standardised documents and responding to common queries.
Implementation Requires Strong Foundations
The Data and Technology track on Tuesday afternoon emphasised that successful AI implementation depends critically on well-organised, accessible data. The adage of “garbage in, garbage out” was shared by many - without proper data foundations, even the most sophisticated AI tools will underperform. Many firms reported current investments in data lakes and warehouses to support their AI initiatives by getting high quality data foundations in place.
“Buy vs. build’ was another challenge that featured - speakers recommended that companies conduct thorough inventories of their current AI applications and carefully assess whether to build custom solutions or purchase existing products based on their specific needs and capabilities.
People and expertise are fundamental for the successful deployment of AI solutions. The debate over AI versus human involvement featured prominently in many panels and was a focus of the AI Summit’s final panel. The consensus from this debate appeared to be a balanced approach: AI can create efficiencies, but human validation and decision-making remain essential. Additionally, it is crucial to consider the extent of AI usage; junior employees must comprehend the activities performed by AI "workers" to build expertise along with their strong professional knowledge and adequately review those tasks when they advance to senior roles requiring significant evaluation responsibilities. Organizations should actively support this dual capability development through targeted training, mentorship, and exposure to AI-driven projects.
Our key takeaway was that firms are optimistic about AI’s significant near-term potential and clear that they must implement AI with appropriate controls and robust data governance to achieve meaningful, sustainable results in this rapidly evolving field.
