FS-ISAC guides on future state of GenAI in financial services

Artificial Intelligence

Global cyber intelligence-sharing consortium, the FS-ISAC, has released a new guide to help financial services firms chart the future evolution and potential risks of generative artificial intelligence (GenAI) technologies within their businesses.

The paper provides an “all-hazards” approach to assessing future AI system risk, with guidance to help firms proactively identify opportunities to integrate AI, whilst also understanding and managing corresponding risks.

“Clarifying these will help financial firms better leverage the technology to maximise business impact and minimise risks,” the FS-ISAC wrote.

The guidance also assists firms with predicting challenges and identifying uncertainties involved in the phases of AI solution implementation.

Among these include a recommendation for firms to define a tolerance threshold for mistakes made by AI systems (for example, the risks of GenAI hallucinations), as well as a need to clarify who is the responsible party when systems go awry.

The paper also stresses the need to confirm the accuracy of AI systems; enabling employees to test outputs for accuracy will be critical as more and more reliance is placed in these systems, with “a clear relationship between AI usage, oversight, and review… essential to healthy productivity and growth, especially in mission-critical scenarios”, the FS-ISAC wrote.

Among the overarching challenges highlighted in the report include the increasing need for AI-supporting infrastructure, the evolving regulatory landscape, copyright issues, and a shortage of staff skilled in AI systems.

In addition, the paper provides realistic short-, medium-, and long-term risk scenarios to help leaders project the outcomes of their choices.

“AI has the ability to completely transform how we do business, but the impact of that transformation largely remains to be seen,” said Mike Silverman, FS-ISAC’s chief strategy & innovation officer.

“This guidance provides an ‘all-hazards’ approach for firms to thoughtfully manage AI implementations, supporting the resilience of the financial sector and helping to safeguard trust in the global financial system in coming years.”