Asset management AI is delivering more visible gains in efficiency than in investor returns. Only 8% of firms in Mercer's February 2026 survey reported measurable improvements in investment returns from AI, compared with 69% reporting better operational efficiency. The findings were cited in Morgan Stanley and Oliver Wyman's September 2026 asset management BluePaper.
Key Takeaways
- AI adoption is widespread, with 55% of surveyed managers using it in at least one investment process.
- Measurable benefits are concentrated in efficiency and research insights rather than investment returns.
- Data quality and compliance concerns remain significant barriers to broader adoption.
Table of Contents
- Asset management AI shows an uneven payoff
- Research adoption outpaces portfolio decision-making
- Data and compliance constrain broader deployment
- Conclusion
Asset management AI shows an uneven payoff
Mercer's 2026 asset management AI survey covered 131 asset managers globally and was published on May 21, 2026. Alongside the 55% that had integrated AI into at least one investment process, another 27% were using it in pilots or proof-of-concept projects.
The clearest reported benefits were operational: 55% cited faster or higher-quality insights, while only 24% reported reduced costs. That distinction matters because faster work does not necessarily produce lower spending or stronger portfolio performance.
The figures describe firms' reported experience, not the size of any return improvement. They also do not establish that AI consistently generates returns above market benchmarks. The Morgan Stanley–Oliver Wyman report said that whether broader research coverage translates into persistent above-benchmark returns remains unproven.
| Item | Details |
|---|---|
| Investment research | 49% of firms had integrated AI. |
| Data processing | 40% of firms had integrated AI. |
| Portfolio construction | 13% of firms had integrated AI. |
| Trade execution | 7% of firms had integrated AI. |
Research adoption outpaces portfolio decision-making
AI use remains concentrated earlier in the investment process, particularly in research and data processing. Adoption is much lower in portfolio construction and trade execution, where investment decisions are implemented.
Only 5% of firms in Mercer's survey gave AI autonomous or semi-autonomous authority over investment recommendations or trades. The technology is therefore largely supporting human judgment rather than replacing it.
This pattern helps distinguish adoption from investment impact: a manager can use AI extensively to organize information without delegating portfolio decisions or demonstrating better returns.
Data and compliance constrain broader deployment
Mercer found that 69% of firms identified data quality or access as a material barrier to broader AI adoption, while 59% cited regulatory or compliance concerns.
Beverley Sharp, Mercer's Global Manager Research Leader, said addressing data, regulatory and integration challenges would be essential to realizing AI's broader potential in portfolio construction and execution.
Despite those constraints, 91% of respondents said they planned to increase their use of AI over the following 12 months. More than half, or 56%, already had a dedicated platform or program to train employees in using the technology.
Conclusion
The survey points to a clear distinction between adopting AI and proving an investment payoff. Managers' stated expansion plans cover the 12 months following the February 2026 survey; whether that wider deployment produces measurable return gains remains unresolved.
Read More News
Disclaimer
This FinanceVelix article provides general information, not personalized financial, investment, legal, or tax advice. Market prices, interest rates and economic conditions can change; investments carry a risk of loss. Facts reflect the material available when published and may change. Check important information independently and consult a qualified professional when needed. Mentioning an asset, company, or service is not an endorsement.
