2. AI is moving rapidly toward the center of investing.

Only one asset owner in the study describes AI as foundational to its technology infrastructure for portfolio analysis and management today. Seven in 10 expect it to be within two years. The study’s assessment of AI maturity reinforces how much ground remains to be covered. On a five-level scale derived from participating asset owners’ responses, most fall between levels two and three, suggesting that organizational readiness for AI remains relatively early even as ambitions accelerate.

Integration between investment and technology teams is similarly limited. Only one asset owner in the study describes its technology specialists and investment professionals as fully integrated, while roughly half say those teams are only partially integrated. That gap matters because asset owners increasingly see the value of AI coming not simply from adding new tools but from changing how investment work gets done. High-quality data, coordinated workflows and connections between human expertise and technology are emerging as prerequisites for deploying AI more deeply across investment organizations.

“The truly courageous organizations are not layering AI onto existing processes. They are reimagining those processes from the ground up around new tools and capabilities.” — Asset owner

3. AI may rebalance investment work rather than replace people.

Asset owners in the study say they expect technology and AI to account for 30% of investment outcomes within three to five years, up from 17% today. They expect the contribution from people to fall to 40% from 53%, while process remains steady at 30%.

The findings, however, point to a change in what people do in the investment process. As AI, data platforms and analytics take on more information gathering, research, monitoring and decision preparation, human expertise would increasingly concentrate on interpretation, judgment, governance and accountability. That includes asking questions, testing ideas and challenging AI-generated outputs.

The model that’s emerging may look less like a transfer of investment decision-making from people to machines than a rebalancing of the relationship between them. More than half of asset owners in the study identify judgment and decision quality as a future priority, while nearly two thirds identify critical thinking as a priority.

That emphasis is reflected in what asset owners want from AI next. Asked where they most want AI capabilities to improve over the next one to two years, asset owners put investment decision-making first, followed by priorities that include efficiency and automation, data integration and real-time insight, consistency and transparency, and risk management and governance.

The ambition extends beyond automating existing work. The larger opportunity may be to amplify what investment professionals can do, the study suggests. AI could enable teams to process more information, extend their analysis across a broader portfolio and learn more quickly, while leaving judgment and accountability with people. For large asset owners, scale may increasingly be measured not only by assets under management, but by how effectively they can amplify their investment capabilities.

4. Adoption is outpacing measurement and oversight.

AI is already producing results. Four in five asset owners report at least some benefits from the technology, including one in 10 who report significant benefits. The gains, however, remain relatively modest. About a third estimate that AI has improved organizational effectiveness by 5% to 10%, while a similar share say it is too early to tell.

Measuring those gains remains difficult. Only one asset owner in the study reports using agreed-upon metrics to measure AI’s impact. For most of the rest, metrics are either still evolving or AI’s impact is discussed without really being measured.

Governance is playing catch-up as well. Only about a quarter of asset owners have made AI a standing board or investment-committee agenda item. Most of the rest review it occasionally at board or committee level, with a smaller group reporting management oversight below the board.

“The primary risks are not technical. They are organizational: governance, oversight and adoption.” — Asset owner

The gap between adoption and oversight could become more consequential as AI moves closer to investment decisions. Nearly two-thirds of asset owners in the study cite both AI errors and hallucinations, as well as overreliance on AI as key risks to manage over the next two to three years. Cybersecurity, confidentiality and data leakage are also among the risks cited.

5. The gains could be substantial — and uneven.

Nearly half of asset owners in the study expect AI to produce large net gains for the asset-owner industry over the next five to 10 years. Four in 10 expect a more uneven outcome, with gains at some asset owners offset by losses at others.

What separates those outcomes may depend less on access to increasingly powerful AI than on how effectively asset owners build around it. Across the study, the same ingredients recur: better-integrated data, redesigned workflows, connections between investment and technology teams, effective oversight and human judgment. Realizing AI’s potential may require sustained investment in these capabilities, together with clearer ways to measure the value they create.

The Thinking Ahead Institute–MSCI Institute Global Asset Owner Technology Peer Study 2026 included 20 global asset owners with a combined USD 5 trillion in assets under management. Participants included five asset owners in the Americas, nine in Asia-Pacific and six in EMEA. The research included survey questionnaires, 18 hours of one-to-one conversations with key fund leaders and 12 hours of senior workshop discussions. The number of respondents to individual questions may vary.