This week’s stories reveal AI governance hardening into institutional infrastructure: the US Treasury’s sector-specific risk framework, the Cambridge/BIS landmark global adoption report, and the EU AI Act’s imminent August deadline are converging to force finance firms from pilots into production-grade compliance. Meanwhile, a deepfake-driven arms race is reshaping insurance fraud detection, and Orion’s ISO 42001 certification signals that third-party AI trust standards are becoming a competitive differentiator in wealthtech.
Top story: Cambridge’s landmark global study — backed by BIS, IMF and WEF — finds 81% of financial firms now use AI but fintechs are three times more likely than incumbents to have reached full transformation, exposing a widening execution gap across the industry.
Cambridge-BIS-IMF Report: Finance AI Hits 81% Adoption But Execution Gap Widens
Cambridge Judge Business School / CCAF · Strategy
The most comprehensive global study of AI in financial services — covering 628 institutions, vendors and regulators across 151 jurisdictions — finds that 81% of firms now deploy AI at some level, but only 14% see it as transformational to their strategy. Fintechs are more than three times as likely as traditional banks to have reached the ‘Transforming’ stage, while 48% of regulators are still in the ‘Exploring’ phase or not engaged at all. The report also flags data privacy as the top AI risk (cited by 73% of all respondents), with a troubling perception gap: AI vendors significantly underestimate adversarial threats compared to the institutions and regulators they serve.
US Treasury Publishes First Sector-Specific AI Risk Framework for Banks
U.S. Department of the Treasury · Regulation
The US Treasury has released two landmark resources for the financial sector: a shared AI Lexicon to standardise terminology across regulatory, legal and technical teams, and a Financial Services AI Risk Management Framework (FS AI RMF) that adapts the NIST AI RMF into 230 concrete control objectives mapped across the full AI lifecycle. Developed through public-private collaboration involving more than 100 financial institutions, the framework is designed to be scalable from community banks to multinationals. While currently voluntary, legal experts warn it will rapidly shape examiner expectations, internal audit standards and AI vendor contract negotiations across the industry.
EU AI Act August Deadline: Banks and Fintechs Face Fines Up to €35M
Cooley Finsights · Regulation
The EU AI Act’s critical high-risk compliance deadline for financial services falls on 2 August 2026, putting credit scoring, insurance underwriting AI, and AML risk profiling systems squarely in scope for fines of up to €35 million or 7% of global turnover. The European Banking Authority has confirmed that the majority of AI use cases at supervised institutions fall into the high-risk category — making this far from an edge case. UK-based fintechs serving EU customers are not exempt, and national competent authorities are expected to request AI Act compliance evidence as part of standard SREP supervisory cycles from 2026 onwards.
Deepfake Fraud Projected to Rise 160%: Insurers Deploy AI to Fight Back
PYMNTS.com · Risk
Deepfake-related fraud incidents — particularly synthetic voice and identity attacks — are projected to surge more than 160% this year, with insurers seeing a 19% rise in synthetic voice fraud in call-centre interactions alone. The Insurance Council of Australia is building a national AI-powered fraud detection platform with EXL and Shift Technology, designed to allow claims flagged by one insurer to surface coordinated fraud patterns across the entire market. The AI arms race is now fully bilateral: the same generative tools enabling fraud at scale are being deployed as the primary countermeasure, with SAS building agentic pipelines that combine computer vision, OCR and LLM reasoning to screen manipulated images before they reach claims decisions.
https://www.pymnts.com/artificial-intelligence-2/2026/ai-arms-race-pits-insurers-against-fraudsters/
Orion Earns ISO 42001 AI Certification — First Known Portfolio Accounting Provider
WealthTech Strategy · Tools
Wealthtech platform Orion has become the first known portfolio accounting provider to earn ISO/IEC 42001 certification for its AI management systems following an independent audit, validating the governance controls behind its Denali AI layer. The move is strategically timed: Orion’s own advisor survey shows most advisors still lack advanced AI tools, positioning the certification as a trust signal at the precise moment adoption is accelerating. The development echoes Outseer’s recent ISO 42001 win in banking fraud prevention, suggesting AI management certification is quickly becoming a baseline expectation rather than a differentiator across regulated financial services.
https://www.wealthtechstrategy.com/post/the-wealthtech-safari-week-of-july-3-2026