Global study reveals asset managers are concerned about the risks that AI introduces

Given the increasing number of asset managers using AI to support investment decision making, 64% of those participating in the research say they are worried about model/algorithm transparency, explainability, and bias. In addition, 62% are concerned about hallucinations where AI generates plausible sounding but false information and presents it as true.

More than half (55%) of global asset managers say they are concerned about the regulatory risks of AI, with 30% saying they are very concerned. A similar amount (58%) of asset managers are worried about the financial risk from AI, including management of credit, market, and fraud risks. Finally, 64% say they are worried about the implementation costs of AI.

To help mitigate AI-enabled risks, nearly all (93%) of asset managers surveyed agree that the best approach is to balance human expertise with technology, using AI to augment or empower risk teams rather than replace them.

Souvik Das, CTO at Clearwater Analytics, said: "What struck me most in this research is how often the same root cause appears, no matter which risk we asked about. Underneath the concerns about skills, culture, governance, and compliance sits one common thread. Firms don't yet fully trust the data feeding their AI. That's a natural stage for any technology this new to move through. The firms navigating it well are the ones treating their data with the same care they bring to the technology itself. That's what turns AI from something people double-check into something they can rely on to tell them the truth about the risk in front of them."