重塑资产生态体系：人工智能在资产管理与资产服务中的角色

Transforming theAsset Ecosystem:The Role of AI in Transforming theAsset Ecosystem:The Role of AI in Table of Content Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Research Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6The Need for AI in Asset Management and Asset Servicing. . . . . . . . . .8Current State of Asset Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 Executive Summary What does the future look like for asset The asset management and servicing landscape is becoming increasingly competitive as traditional players facegrowing pressure from innovative fintech startups that are redefining how financial services are delivered. Facedwith sustained margin pressure, complex and evolving regulations (e.g., EU Retail Investment Strategy, MiFIDII, DORA, EU AI Act, …), as well as rising client expectations for transparency, customized services and digital In this context, Artificial Intelligence (AI) can no longer be considered a breakthrough concept. It is becoming anessential lever for transformation across the asset management value chain — from portfolio decision support tooperational efficiency and compliance. AI, as defined in this paper, refers to systems capable of performing tasks Yet, while the potential of AI is widely recognized — 92% of organizations acknowledge the strategic importanceof digitization and new technology integration — only 52% report partial AI integration, and most remain at thepilot stage. The gap between ambition and reality is not just technical: it reflects deeper challenges related to datareadiness, regulatory complexity, and organizational adaptation. This observation is further confirmed by the 2024AI thematic report from the CSSF, which highlights that although AI adoption is progressing, a significant portion Our findings reveal that AI is primarily used today to augment rather than replace human decision-making. 78% offirms do not use AI for real-time decisions, and instead focus on support functions that enhance judgment, reduce Despite this conservative deployment, early results are promising: •69% of respondents report reduced manual processing time in pre-trade activities;•55% note improved decision quality; However, the road to broader implementation is still marked by three main constraints: 1.Data quality and infrastructure limitations (only 9% of firms have fully standardized and accessible data);2.Limited internal expertise; Many leading players are responding by adopting hybrid models, owning the strategy and expertise whileoutsourcing technology components, particularly through partnerships with AI vendors. Commonly outsourcedelements include pre-trained AI models for document analysis, cloud-based infrastructure for model deployment,APIs for generative AI, or intelligent automation solutions for processing unstructured data. This allows Maintaining in-house understanding of AI capabilities and their implications allows firms to steer technologyadoption in line with their operational models, regulatory constraints (notably EU AI Act), and long-term strategic Looking ahead, the success of AI transformation will depend less on technology itself than on the ability tointegrate it meaningfully within the organization. Research confirms that 60% of transformation success depends Firms that succeed will be those that position AI as a collaborative co-pilot — one that amplifies, rather thanreplaces, human expertise — and embed it into their operating models in a way that is scalable, explainable, andaligned with regulatory and ethical expectations. This is, however, only a first step. The final stage is to evolve froman “augmented employee” to an “augmented organization” where the operating model is transformed by putting Asset Management -In the context of this paper, Asset Management encompasses both core investment activities(such as portfolio construction and fund selection) and the operational and administrative functions commonly Research This white paper examines AI adoption in asset management by assessingorganizations’ maturity, collecting executives’ perspectives of the Europeanmarket and identifying the prerequisites for becoming an augmented Data Collection The primary data is gathered through a series of semi-structured interview calls conducted between Octoberand March 2025. The sample includes more than 40executives holding key positions (CEO, COO, CTO, Head The participating organizations also differ significantlyin size, from specialized firms with fewer than 100employees to large global institutions with morethan 10,000. This diversity offers a comprehensive The interviews focus on organizations operating withinEurope. with a majority based in Luxembourg, reflectingits role as a key hub for asset management andservicing. Insights are also gathered from participants The participatin