Asset managers have moved well past the exploratory phase of artificial intelligence adoption, but they are keeping the technology at arm's length when it comes to the decisions that actually move markets.
That is the central tension in Mercer's 2026 AI in Asset Management Survey, which found that 55% of asset managers report AI is integrated in at least one of their strategy's investment processes, 27% are at pilot or proof-of-concept, and only 18% report no integration yet. The survey drew on responses from 131 managers around the world, gathered in February and March 2026.
The numbers reflect an industry that has broadly embraced AI as a tool for doing existing work faster and more thoroughly but has stopped well short of handing it genuine autonomy.
Most firms describe their AI integration as operational as 73% use it for automation and efficiency or in a "co-pilot" capacity, while 68% deploy it for insight and analysis, and only 5% said AI is used for autonomous or semi-autonomous decision-making for investment recommendations or trades.
"AI is delivering measurable efficiency and insight for asset managers today, but the technology is largely a partner rather than a decision-maker," said Beverley Sharp, Mercer's Global Manager Research Leader. "Addressing data, regulatory, and integration challenges will be essential to realize AI's broader potential in portfolio construction and execution."
The practical results of that approach show up in the reported benefits. The most common measurable gains were enhanced operational efficiency, cited by 69% of respondents, and faster or higher-quality insights and decision-making, noted by 55%. Improved returns and reduced risk or volatility were each cited by just 8%.
Where precisely in the investment workflow AI lives matters enormously for allocators trying to separate substance from salesmanship. Today's most common already-integrated use cases are idea generation and research, processing unstructured and external datasets, and signal generation and market trend analysis. Far fewer asset managers report AI embedded in portfolio construction or trade execution.
The barriers keeping AI upstream are not primarily philosophical. Data constraints and regulatory concerns are the dominant frictions, with 69% of firms citing data quality or access issues and 59% pointing to regulatory or compliance concerns as material obstacles. Firms also expressed worry that existing regulatory frameworks may not adequately cover certain aspects of AI use — 31% flagged data governance gaps and 24% identified system-level risks, such as herding behavior, as the most significant blind spots.
Phrases like "we use AI" or "powered by AI" can mean anything from rudimentary automation to production-grade models actively influencing live portfolios, and effective due diligence should determine whether a manager's use of AI is purposeful, credible, well governed, empirically validated, and sustainable at scale.
For allocators conducting that due diligence, the right questions center on what specific investment problem the manager is trying to solve, what tasks AI actually performs, and how those systems are governed. Strong answers articulate a clear hypothesis for how AI improves outcomes — whether through signal discovery, broader coverage, cost reduction, or risk management — and specify where in the process AI operates.
Firms appear to recognize that realizing AI's potential requires internal cultural shifts alongside technological ones. More than half — 56% — have a dedicated platform or program in place to educate employees on how to use AI within their roles.
"Our survey finds AI is mainly being used upstream for idea generation and research in the asset management industry. This mirrors our experience with AI," Sharp said. "Across Mercer, we're innovating, using the technology to help our clients, including developing an AI-powered manager research tool to help streamline the collation of data and the drafting of due diligence documents, with more tools to come."
When managers tout AI-driven outperformance, calibrating what that claim actually means is worth the effort. The survey's breakdown of reported benefits suggests that in most cases, the gains are concentrated in operational speed and research quality rather than in the performance statistics that matter to end investors.
The pace of change is unlikely to slow. Almost all surveyed managers — 91% — said they plan to increase their use of AI in the next 12 months, making it a fixture of manager due diligence for the foreseeable future. Whether that expansion pushes further into portfolio construction and trading — or remains confined to the research layer — will define the next chapter of AI's role in asset management.
Two firms land teams managing more than $1.1 billion in combined assets from Kestra and Edward Jones.
A private partnership, Edward Jones is a giant in the retail brokerage industry with more than 20,000 financial advisors.
Meanwhile, Raymond James and Tritonpoint Partners separately welcomed father-son teams, including a breakaway from UBS in Missouri.
Paul Atkins has asked staff to solicit public comment on novel ETFs, pausing the clock on as many as 24 filings linked to the booming event contracts market.
From 401(k)s to retail funds, Deloitte sees private equity and credit crossing into mainstream investing on two fronts at once.
Wellington explores how multi strategy hedge funds may enhance diversification
As technical expertise becomes increasingly commoditized, advisors who can integrate strategy, relationships, and specialized expertise into a cohesive client experience will define the next era of wealth management
