This week, the finance and insurance sectors are navigating a convergence of pressures: the EU AI Act’s high-risk deadline has just landed, insurers are scrambling to price AI-related liabilities they barely understand, and Revolut’s proprietary AI brain is raising the bar for what a vertically integrated financial AI model can do. Across banking, buy-side investing, and insurance, AI is shifting from pilot to infrastructure — but governance, licensing, and coverage gaps are exposing serious institutional blind spots.
Top story: Revolut’s proprietary foundation model PRAGMA boosts fraud detection by 64.7% and credit risk prediction by 16%, setting a new benchmark for vertically integrated banking AI.
Revolut’s PRAGMA Model Redefines What Bank AI Can Do
Forbes · Strategy
Revolut has built PRAGMA, a proprietary AI foundation model that unifies all customer interactions — transactions, app usage, investments, and support — into a single connected system. The model delivers a 64.7% improvement in fraud detection and a 16% uplift in credit risk prediction, powered by NVIDIA GPUs at a scale of 70 million users. Unlike rivals stitching together fragmented third-party tools, PRAGMA signals a new competitive moat: financial institutions that own their AI stack may increasingly outperform those that don’t.
Insurers Begin Excluding Autonomous AI From Cyber Policies
Fintech Global · Risk
A new fault line is opening in the cyber insurance market: insurers are tightening policy wording and introducing AI-specific exclusions where autonomous AI risk is deemed too difficult to quantify or price. A Delinea survey found that 42% of companies now have AI-related exclusions in their cyber policies, while firms that deploy AI defensively may receive premium discounts. For financial services firms renewing cover in 2026, the question is no longer whether AI is covered — it’s which AI, under which conditions, and in which policy.
https://fintech.global/2026/07/28/why-autonomous-ai-could-void-your-cyber-insurance-in-2026/
EU AI Act High-Risk Rules Now Live: Banks and Insurers on the Hook
EVE Core / Finextra · Regulation
As of 2 August 2026, the EU AI Act’s high-risk obligations became fully enforceable, directly capturing credit scoring, insurance risk assessment, fraud detection, and AML systems used by banks and insurers across the EU. Fines of up to €30 million apply for non-compliance, and crucially, deployers cannot outsource their obligations to AI vendors — contracts must be restructured to include audit rights and model-change notifications, most of which currently contain none. UK firms with EU customers or EU-regulated entities are equally exposed and cannot treat this as someone else’s problem.
Buy-Side AI Blocked by Broker Research Licensing Bottleneck
Traders Magazine / LLRX · Tools
A new Substantive Research and Aiera survey reveals that 77% of buy-side firms have organisation-wide generative AI deployments in place, yet the same proportion cite broker research as the most valuable data they want as machine-readable feeds into their AI systems — and licensing restrictions are preventing it. The finding exposes a structural friction point: the buy side has built the AI infrastructure but cannot legally feed it the data that matters most. For asset managers, solving broker research licensing is now as strategically important as the AI build itself.
https://www.llrx.com/2026/07/ai-in-finance-and-banking-july-16-2026/
New Study: AI Co-Advisors Deliver Scalable, Affordable Financial Advice
HackerNoon · Finance
A peer-reviewed study published in July 2026 — ‘Financial advice behaviour: humans versus AI’ — concludes that AI co-advisors can deliver more scalable and affordable financial guidance than human-only models, adding academic weight to the rapid growth of AI-driven wealth management. The findings land at a critical moment: McKinsey projects nearly 40% of financial advisors will retire within a decade, creating a shortfall of roughly 100,000 professionals, while $22 trillion in assets is set to transfer to digitally native Gen X and millennial investors. For wealth managers and robo-advisory platforms, the research case for AI-augmented advice is now stronger than ever.
https://hackernoon.com/has-ai-in-fintech-become-powerful-enough-to-replace-financial-advisors