How Early Asset Management Adopters Are Redefining Research, Operations, and Competitive Advantage
A CFA Institute 2024 survey report revealed that artificial intelligence (AI) was the most raised issue in the past 12 months among asset manager respondents, eclipsing every issue from private markets to cryptocurrency.i If you are worrying you are falling behind in the race to operationalize AI, you’re not alone. Capital markets investment managers are notoriously resistant to technological change. Just yesterday, it seemed the industry was wrapping its brain around rudimentary AI functions like customer service chatbots and document summarizations. Now, AI investment priorities are moving beyond the middle- and back-office to include a range of front-office functions. Many prefer to learn from the innovators and early adopters' mistakes before joining the majority, especially as two-thirds of early adopters would revise their initial strategies.ii However, the AI revolution may be much more monumental than previous shiny new objects.
McKinsey asserts that adoption could impact asset managers to the tune of 25-40% of their cost base.iii To say such substantial ROI is transformative is an understatement. But what have early adopters of AI really done, and how much impact have they made? As of now, early AI adopters among institutional investors and asset managers have achieved a certain level of competency in two main categories: research and investment operations.
Where early AI adopters are creating the most impact: research
Research-related areas are where most early adoption activity is happening within the asset management community, focused on use cases like aggregating research reports, analyzing new datasets, and portfolio construction.
Earnings analysis
Firms are using AI to analyze earnings and earnings calls in a routine way. Analysts and portfolio managers (PMs) have a lot of quarterly earnings calls to listen to. Today, AI tools can instantly review hundreds of earnings call transcripts from platforms like Bloomberg Terminal, Capital IQ, and FactSet data feeds. Subsequently, if the firm’s AI has natural language capabilities, analysts can query the AI in plain English to provide a summary, appraise the tone, flag relevant changes, and ask specific questions. This aids security and industry analysis, especially the incremental changes or deltas.
Alternative data analysis
The analysis of alternative datasets, from credit card data and web traffic to macroeconomic reports, is another use case for asset managers. Mountains of alternative data are great to have but not so easy to curate to distill insights. AI is stepping forward to help narrow the focus of research from reams of disparate information. Moreover, firms are using natural language AI tools to interrogate datasets and build custom reports and analyses, without the need to chase other departments or IT to access that information.
Portfolio construction and risk analysis
Portfolio and risk managers used to huddle up and do traditional market and fundamental analysis, position sizing over spreadsheets, and run exposure analysis and scenario testing. More recently, they have turned to AI to succinctly advise on whether they are underweight or overweight in specific positions and on whether they should adjust based on value-at-risk (VaR). Managers are using AI become speedier to insights so they can adjust their portfolios in response to market shifts “by establishing a total portfolio view, increasing visibility into the fund’s exposure to risk factors and performance drivers.”iv
Large asset managers have built internal platforms to process these large data sets for research, go through them on a daily or weekly basis, and generate trading signals. However, we are talking about a ton of unstructured datasets, which, if not properly ingested and normalized into the firm’s data platform, will render AI tools hamstrung. Some systems are ill equipped to handle information within PDFs, including LP/GP statements, capital calls, emails, and, if trading in private debt, agonizing long loan tapes. Before AI can help PMs analyze unstructured data to drive alpha, AI tools must extract and structure this information properly, reducing manual review time and improving accuracy.
The next step in AI strategy: investment operations
Early adopters are proficient at using AI for research. Further, the ROI for this function is obvious: alpha generation. Firms are now applying learnings from these research-oriented AI projects to investment operations, a use category still in the very early stages of adoption. The ROI for investment operations is primarily efficiency — handling more portfolios and lowering risks. But the ROI is also about making investment data a true strategic asset, beginning with shoring up data quality and governance.
Data connectivity and curation
Managers are leveraging AI to simplify data connectivity, improve data quality, understand new datasets faster, and build usable data products. AI is adept at automating critical data management functions like preparing, processing, and normalizing new incoming financial data. Then, firms are using AI to locate, explore, and understand their datasets, for advanced analysis, visualizations, and reporting. Firms that have self-service, low-code AI and data platforms enable their people to execute the above functions without technical know-how, only using natural language prompts.
An AI reconciliation revolution
Managers are experimenting with leveraging AI for more automated, pattern-driven reconciliation and managing corporate actions. Reconciliation is a taxing manual process required to synchronize information stored across disparate systems that may hold different versions of the same data. AI can help cut out time-intensive manual human intervention, allowing a firm’s skilled humans to set their sights on higher-value tasks.
Corporate actions processing
Inadequate operations technology and poor operational oversight can lead to a manager thinking they have a different position than they truly hold, primarily due to a missed or improperly booked corporate action. If a firm handles corporate actions manually and is a high-volume business, it may necessitate hiring many people just to process trades. As volumes skyrocket, this is becoming less sustainable.
AI can help handle automated lifecycle events, including corporate actions on public equities or mergers and acquisitions. This ensures that when PMs arrive in the morning, the corporate actions have already been auto processed, empowering the front office with the right information promptly. AI-powered automation for investment operations like reconciliation and corporate action processing is essential in today’s complex markets.
Solving the unstructured data problem in public and private markets
Unstructured data presents thorny problems for operations as well as research. The management of capital calls is one critical area. Without automation in place to speed the tracking and generation of capital calls, the manual effort is onerous and error prone. It requires utter precision to ensure critical transactions are executed on time. For example, institutional investors like pension funds are increasingly seeking AI-powered automation because they generally trade in 75% public securities and 25% private asset classes. GPs send LPs quarterly, weekly, and in some cases daily communications for things like capital calls on these unstructured products in the form of security information and transaction information. And it still comes in PDF and spreadsheet formats. LPs literally have people go through 100 or 200 documents and manually input them into an Excel sheet. Without AI, it's simply impossible to deal with this volume of frequent, scattershot data.
What is your AI readiness?
AI adoption in asset management for research is headed for the mainstream. Early adopters are now aggressively tackling use cases for investment operations, and are considering the ROI of these efficiencies. Large wealth managers are significantly scaling, with 35% planning more than 15 new use cases within 2 years. More front-office use cases will emerge as AI proliferates. Firms with a modern data foundation in place will undoubtedly be nimbler in moving from tactical AI integration to more strategic integration.v
That foundation should be data-first, cloud-ready, and tightly governed — a combination that enables AI to scale responsibly. Vendor platforms are crucial for incorporating AI into research and operations as issues like regulatory compliance, data privacy, and scalability are too much for firms to handle accurately on their own. The cost of inaction now outweighs the risk of modernization. All firms should ask the question, “What is my level of AI readiness today, and does it prepare me for tomorrow?”
Authored By
Mahesh Narayan
Mahesh oversees Arcesium’s capabilities for institutional asset managers, including the Arcesium Data Platform, middle- to back-office solutions, and associated financial operations.
Share This post
[i] CFA Institute, October 2024. https://blogs.cfainstitute.org/investor/2024/10/11/an-industry-in-transition-ai-top-of-mind-in-2024-asset-manager-survey/
[ii] EY, September 16, 2025. https://www.ey.com/en_us/insights/wealth-asset-management/gen-ai-in-wealth-asset-management-survey
[iii] McKinsey, July 16, 2025. https://www.mckinsey.com/industries/financial-services/our-insights/how-ai-could-reshape-the-economics-of-the-asset-management-industry?
[iv] McKinsey, January 21, 2025. www.mckinsey.com/industries/private-capital/our-insights/unlocking-value-from-technology-and-ai-for-institutional-investors
[v] EY, September 16, 2025. https://www.ey.com/en_us/insights/wealth-asset-management/gen-ai-in-wealth-asset-management-survey
