Finance firms face USD 14 million average cost per conduct incident as AI risks surge
Financial institutions are experiencing conduct risk incidents at accelerating rates, with costs averaging USD 14 million per incident. Major incidents rose 55% between 2023 and 2025, according to analysis from RepRisk released today, based on a survey of more than 500 C-suite executives across banks, asset managers, and other financial firms.
Firms face between USD 28 million to USD 43 million in annual cost exposure from reputational and business conduct risks, experiencing two to three significant incidents per year on average. The most severe incidents cost USD 37.6 million.
AI risks now top the list of material threats
The share of executives identifying AI-related conduct risks as a top material risk jumped sharply. Only 16% flagged AI conduct risks as material over the past three years, but 56% expect them to rank first among non-financial risks over the next three years.
The concern reflects how deeply financial firms are embedding AI into core workflows-from transaction due diligence and risk monitoring to portfolio oversight, compliance, and KYC processes. At scale, flawed data can spread errors across models, dashboards, portfolios, and decisions in ways that become difficult to identify or reverse.
Overall risk complexity has increased for 67% of executives surveyed over the past year.
Financial firms prefer human oversight in AI systems
When AI-driven decisions inform material risks, finance executives show clear preference for human-led approaches. Across the sample, 73% use human-AI hybrid models, and 67% trust hybrid data for material risk and investment decisions, compared with 35% for AI-only approaches.
Banks show even stronger confidence in hybrid systems. Seventy-four percent of bank respondents trust human-AI hybrid data, signaling demand for technology enhanced by expert oversight rather than full automation.
Early detection could cut losses significantly
Even modest improvements in monitoring could yield substantial savings. Reducing incident frequency by 5% to 10% or accelerating escalation before issues intensify could mitigate multi-million-dollar losses annually.
Most financial firms still invest in conduct data reactively rather than preventively. Executives surveyed expect return on investment from structured conduct risk intelligence to double within three years.
For finance professionals managing these risks, understanding both the financial exposure and the operational limits of AI systems is now essential. Learn more about AI for Finance or explore the AI Learning Path for CFOs to build expertise in this area.
Your membership also unlocks:
