金融业人工智能实践指南

I N S I G H TR E P O R T Contents Foreword Executive summary Introduction 1The landscape today 1.1Context and industry outlook1.2AI spending trends and prioritization 2Investment and maturity of the industry 2.1Business-driven adoption models2.2Aligning AI to strategy2.3Measuring AI value 3.1The foundation and platform architecture3.2Skills, hiring and culture change3.3Governance, risk and compliance management 4.1Why agentic AI is key4.2Human–AI partnership 5Planning for the industry’s future32 5.1Emerging technologies and cross-industry innovation335.2Framework for scaled and sustained AI adoption335.3Leadership, sponsorship and accountability35 A3Additional reading40 DisclaimerThis document is published by the World Economic Forum as a contribution to a project, insight area or interaction. The findings, interpretations and conclusions expressedherein are a result of a collaborative process facilitated and endorsed by the World ©2026 World Economic Forum. All rights reserved. No part of this publication maybe reproduced or transmitted in any form or by any means, including photocopying and recording, or by any information storage and retrieval system. Foreword David ParkerGlobal Industry Lead,Banking and Capital Markets, Drew PropsonHead, Technology andInnovation in Financial Services, Andre BelelieuHead, Financial ServicesIndustries; Head, Business The artificial intelligence (AI) landscape of 2026 wouldhave seemed implausible just a few years ago,with models that reason across complex problems,agents that autonomously execute multi-stepworkflows, and capabilities that are fundamentally This report is the outcome of Phase II of theinitiative, which has engaged over 150 seniorleaders from more than 100 organizations. Building on the Forum’s 2025 white paper,Artificial Intelligence in Financial Services, whichexamined AI’s impact on firm-level strategy andthe broader financial ecosystem, this report offerspractical guidance for organizations at everystage of their AI journey. It is grounded in casestudies that illustrate real-world applications withmeasured results, and is enriched by insights from AI has been a priority on the World EconomicForum’s financial services agenda for more thana decade, resulting in a large body of research.Over that time, the conversation has transformed,from the exploration of a technology in its nascent In 2024, recognizing that the latest wave ofAI advancement demanded an intensive andstructured response, the Forum and Accenturecollaborated to launch a multi-phase AI in FinancialServices initiative, bringing together a diverse set We hope this report provides confidence andclarity as AI continues to evolve at a remarkablespeed, and that the insights within, garneredthrough trusted dialogue among peers acrossregions, contribute to a financial sector that Executive summary As artificial intelligence reshapes financialservices at an intense pace, institutions A scalable enterprise intelligence platformis essential. Institutions need a unified layerthat coordinates data, models, decisionsand automation, underpinned by robust datamanagement, identity controls, auditability and Artificial intelligence (AI) represents a structuraldisruption, and its impact on financial servicesgoes far beyond efficiency gains. It is reshaping As the industry moves from experimentation toscaled deployment, the central challenge fororganizations is fully integrating AI quickly andsecurely. The decisions involved, from datamanagement to governance structures, are People must remain firmly in the lead. SuccessfulAI adoption depends on people embracing anddriving the transformation. This will take the form of ahybrid workforce, where people and AI agents worktogether supported by evolving roles, skills-based The report’s key findings are as follows: Risk management and responsible AI need to beembedded at every layer. Institutions must define riskappetite early, validate models rigorously and ensure AIactions remain explainable, traceable and controllable.This means deepening model-risk disciplines and AI must be a strategic leadership issue.The priority has shifted from running pilots toembedding AI across the enterprise, quickly,responsibly and in ways that deliver lastingvalue. Leaders must define where to focus,what to automate and where human judgement A two-speed implementation model sustainsmomentum. Institutions are increasingly pursuing twospeeds simultaneously, using AI to deliver near-termgains in productivity, experience and decision quality,while building the enterprise foundations required Differentiation will increasingly come fromcustomer relationships amplified by AI. Withanalytical and generative capabilities now widelyaccessible, organizations that use AI to intentionallystrengthen customer relationships will be better Although there is no single path to success andrequirements will inevitably evolve, organizationsthat proactively define strateg