AI Exposure Is a Total Portfolio Question
By: Frank Benham and Steve MacLellan, Meketa Investment Group
While some investors may regard equity concentration among U.S. technology companies as simple company-specific risk, others are looking at this phenomenon as one channel of what has become a cross-asset driver: artificial intelligence. This exposure to AI that is prevalent in the U.S. stock market also appears, through different mechanisms, in private equity, private credit, real estate, infrastructure, and fixed income.
Most public pension plans that have reviewed their artificial intelligence (AI) exposure have done so within their public equity allocation. That review is likely incomplete. AI has become a cross-asset driver that appears in private equity, private credit, real estate, infrastructure, and fixed income. A plan that evaluates AI only through its equity sleeve is not seeing the whole picture.
What Counts as AI Exposure?
A tight definition of AI exposure includes only direct participants in the AI value chain, specifically the firms designing and manufacturing computational resources (“compute”).
A moderate definition adds firms that benefit from capital expenditures of AI companies. This includes power generators, grid equipment makers, and specialized data center developers whose revenues depend on the AI buildout continuing. A wide definition further includes indirect beneficiaries and adversely affected parties, such as AI-enabled incumbents whose competitive positioning depends on adoption.
Scope choice is consequential and should be applied consistently. Applying it differently across asset classes produces a picture that is internally inconsistent regardless of whether individual measurements are accurate.
How AI Exposure Enters the Portfolio
In public equity, AI concentration is visible. AI-linked names dominate the list of largest companies in the S&P 500. Under the moderate scope definition described above, approximately 48% of the MSCI USA Index carries AI-linked exposure. For a globally diversified equity portfolio, the figure is lower but still substantial, at approximately 37% of the MSCI ACWI Index.1
Some hyperscalers turned to corporate bond markets at scale beginning in 2025. A plan holding a broad investment grade index now carries hyperscaler credit exposure estimated at roughly 2–4% of Bloomberg Aggregate holdings depending on scope. Private equity exposure depends heavily on vintage and strategy mix. Mature buyout-heavy portfolios may have 10–25% of holdings that are meaningfully AI-adjacent under a moderate scope definition, even if 30–50% is labeled as technology. Recent vintage venture capital is more heavily weighted toward AI, including access to foundation model developers.
Data centers now drive meaningful AI exposure in both real estate and infrastructure. Listed REIT benchmarks carry roughly 10% data center exposure, while infrastructure allocations may have 10–30% in AI-adjacent themes including power generation for data center load. Private credit has grown as a channel for AI financing as well.
The Counterparty Problem
Across these channels, many exposures trace back to the same small set of hyperscalers. These firms appear as equity holdings in the public portfolio, as investment grade issuers in fixed income, as data center tenants in real estate, as offtake counterparties in infrastructure, and as borrowers in private credit. A plan that sizes each position appropriately for individual asset classes may nonetheless be concentrated at the total portfolio level.
What a Total Portfolio View Shows
A total portfolio view shows that while public equities are the most visible AI channel, they are only one part of the portfolio’s connection to AI. Private markets, credit, real estate, and infrastructure can all add exposure through capital spending, financing, data center demand, power needs, and common counterparties. For investors, the more important question is not how much AI appears in the equity allocation alone, but how much cumulative exposure exists across the whole portfolio.
Governance Implications
The goal is not to argue for raising or lowering AI exposure. It is to make the exposure known and deliberate. An investor who measures cumulative AI exposure and concludes it falls within their intended range has made a conscious choice. An investor who has not measured it has simply accepted whatever the portfolio contains. Plans may also want to consider drawdown scenarios. Asset class diversification reduces risk when asset classes respond to different drivers. When a single driver dominates, correlations in a stress event may be higher than standard portfolio construction assumes.
Key questions for boards and investment committees include: What scope defines AI-linked exposure, and is it applied consistently across asset classes? How does cumulative AI exposure fit within the existing risk budget? How concentrated is the plan in a small number of ultimate counterparties across multiple asset classes?
These questions do not require precise answers to be useful. A plan that establishes its cumulative AI exposure is meaningfully larger than its equity allocation alone may have already captured most of the governance value the exercise offers.
This article is adapted from “AI Exposure: Implications Across Asset Classes,” a Meketa Investment Group whitepaper published in May 2026.
Frank Benham, CFA, CAIA, Managing Principal/Director of Research, joined Meketa in 1999 and serves as Director of Research, leading the firm’s thought leadership efforts, including research publications, portfolio construction frameworks, and capital market expectations. He has authored or co-authored numerous papers on asset allocation, risk management, and private markets, helping shape Meketa’s investment philosophy and guidance to institutional clients. He chairs the firm’s Investment Policy and Strategic Asset Allocation/Risk Management Committees and is a member of the Private Markets Policy Committee.
Mr. Benham holds a finance degree from Bentley College and is both a Chartered Financial Analyst® (CFA) and Chartered Alternative Investment Analyst (CAIA). He is a member of the CFA Institute, CFA Society Boston, and CAIA Association®. Prior to Meketa, he worked at State Street Bank. He is also an occasional speaker at industry conferences and institutional forums.
Steve MacLellan, CFA, Managing Principal/Consultant, joined Meketa in 2013 and serves as lead consultant for Taft Hartley and public fund clients, supporting defined benefit pension plans and health and welfare funds. His responsibilities include investment policy design, asset allocation modeling, and evaluating manager and fund performance. He is also a member of Meketa’s Sustainability and 401(k) Plan Committees.
Prior to joining the firm, he was a senior associate in the Fund Oversight group at State Street Global Advisors, where he worked across a broad range of asset classes and investment strategies, from passive indexing to active absolute return mandates.
Mr. MacLellan graduated summa cum laude from the University of Massachusetts Amherst with a degree in economics and political science. He holds the Chartered Financial Analyst® designation and is a member of the CFA Institute and CFA Society Boston.
Endnotes:
1. Exposure for public markets is as of April 30, 2026, and for private markets is as of December 31, 2025.
Disclosures: THIS REPORT (“REPORT”) IS INTENDED SOLELY FOR THE RECIPIENT. INFORMATION, OPINIONS, OR RECOMMENDATIONS REFLECT OUR VIEWS AS OF THE DATE SHOWN AND ARE SUBJECT TO CHANGE. WE HAVE RELIED ON DATA FROM EXTERNAL SOURCES AND MAY HAVE USED ARTIFICIAL INTELLIGENCE (“AI”) TOOLS IN PREPARATION. WHILE CARE HAS BEEN TAKEN, WE CANNOT GUARANTEE THE ACCURACY, COMPLETENESS, OR RELIABILITY OF ANY INFORMATION, WHETHER FROM EXTERNAL PROVIDERS OR AI.
ALL INVESTMENTS INVOLVE RISK, AND THERE IS NO ASSURANCE THAT STRATEGIES DISCUSSED WILL SUCCEED. PAST PERFORMANCE IS NOT INDICATIVE OF FUTURE RESULTS. THE REPORT MAY INCLUDE FORWARD-LOOKING STATEMENTS BASED ON CURRENT ASSUMPTIONS, WHICH ARE SUBJECT TO CHANGE; ACTUAL OUTCOMES MAY DIFFER MATERIALLY. RECIPIENTS SHOULD CONSULT THEIR MEKETA ADVISOR OR ANOTHER PROFESSIONAL BEFORE MAKING FINANCIAL DECISIONS. WE DISCLAIM LIABILITY FOR ANY DAMAGES ARISING FROM USE OF THIS REPORT.