America’s public pension funds have invested hundreds of billions of dollars in AI-related companies. The boom has helped strengthen retirement portfolios, but what happens if the market turns?
A retired teacher in California, a firefighter in Texas, and a municipal employee in New York may have something in common they have never considered: Part of the money supporting their retirement could depend on the financial success of artificial intelligence.
They may never have purchased shares of Nvidia, Microsoft, or Alphabet. They may know little about artificial intelligence beyond the headlines. Yet their pension funds, like many institutional investors, hold substantial investments in the technology companies driving the AI boom.
For years, those investments have helped generate strong returns. But as Wall Street pours money into artificial intelligence, a troubling question is emerging: What happens to retirement security if today’s technological excitement turns into tomorrow’s financial disappointment?
The concern is no longer hypothetical speculation among market analysts. Public pension officials are beginning to examine how much of their financial future depends on AI-related investments.
More Than Half a Trillion Dollars
A study released in October 2026 by Equable Institute, a nonprofit organization that analyzes public retirement systems, estimated that state and local pension funds held at least $578.1 billion in AI-related companies as of the second quarter of the year.
That represented at least 9 percent of their total assets. The estimate covers reported investments in 52 publicly traded AI-related companies. It does not fully account for private equity, private credit, or investments managed through outside firms. The actual exposure could therefore be substantially higher.
This is not necessarily the result of pension managers deliberately gambling on artificial intelligence.
Many pension funds invest in broad stock market indexes. As companies such as Nvidia, Microsoft, Alphabet, Amazon, and Meta have become increasingly influential in financial markets, their shares have also become increasingly important components of retirement portfolios.
In other words, a pension fund can become heavily exposed to AI without ever announcing a strategy to invest specifically in artificial intelligence.
That creates a form of concentration risk: Investments that appear diversified across several companies may still depend on the same technological and economic expectations.
Why Pension Funds Invest in AI
Public pension systems face a difficult financial obligation. They must invest money today to pay benefits that workers may not collect for decades. Their income generally comes from contributions by employees and public employers, together with returns on investments.
To meet long-term commitments, pension managers invest in stocks, bonds, real estate, private businesses, and other assets. Their challenge is to generate sufficient returns without taking risks that could threaten future payments.
AI-related companies have been attractive because investors expect artificial intelligence to transform industries, increase productivity, and generate new sources of revenue.
The optimism has pushed valuations higher, benefiting investors who already owned shares. For pension funds, those gains can improve financial health and reduce pressure on governments to increase contributions. But there is a fundamental difference between a technology that changes the world and an investment that produces the returns investors expect.
The internet transformed commerce, communication, and society. Yet many internet companies collapsed when the dot-com bubble burst in 2000.
A technological revolution can be genuine even when the stock market becomes excessively optimistic about its immediate profitability.
What If the AI Boom Becomes a Bubble?
A financial bubble develops when asset prices rise beyond levels that can reasonably be supported by future earnings or economic fundamentals. Some analysts worry that enthusiasm for artificial intelligence may be encouraging investors to assume extraordinary growth for many years.
Technology companies are spending enormous sums on data centers, advanced chips, electricity infrastructure, and AI research. These investments may eventually produce substantial profits, but their financial returns remain uncertain.
If revenues fail to justify the spending, investors could reconsider the valuations assigned to AI-related companies.
A sharp decline would affect not only individual shareholders but also pension funds, retirement accounts, and other institutions holding those assets.
Consider a simplified example. Imagine a pension fund with $10 billion in assets, of which $1 billion is invested in AI-related companies. If those holdings lose 40 percent of their value, the fund would suffer a $400 million loss from that portion of its portfolio, assuming everything else remains unchanged.
The fund would still own other investments, and market values could eventually recover. But the immediate loss would weaken its financial position.
The more troubling scenario would be a broad market downturn in which technology stocks, corporate bonds, real estate, and other assets decline together.
That is why pension officials are increasingly concerned about the possibility that apparently different investments may share exposure to the same AI-driven economic cycle.
The Warning Signs Are Already Visible
The debate has reached some of America’s largest retirement systems.
The California Public Employees’ Retirement System, known as CalPERS, manages more than $600 billion and serves over two million members. Its investments include major technology companies that have benefited from enthusiasm for AI.
At a September meeting, board members raised questions about whether the financial rewards of technology investing were accompanied by growing downside risks.
In Los Angeles, officials examining a public retirement portfolio found that as much as 19 percent of its investments could be connected to artificial intelligence, depending on how exposure was defined.
New York City pension officials have also expressed concern about excessive concentration. According to reporting by The New York Times, the city’s pension leadership declined an approximately $250 million private equity opportunity involving AI businesses because of concerns about the portfolio’s exposure to the sector.
These decisions do not establish that an AI bubble exists or that a collapse is imminent.
They show that professional investors responsible for workers’ retirement benefits are beginning to question how much risk they should accept from a single technological trend.
Could Retirees Actually Lose Their Pensions?
This is the distinction that matters most to ordinary Americans.
A decline in a pension fund’s investment portfolio does not automatically mean that retired workers will receive smaller pension checks.
Traditional defined-benefit pensions generally promise payments calculated according to factors such as salary, years of service, and retirement age. The payments are not simply the current value of an individual’s investment account.
Many public pensions also operate under legal protections governing promised benefits, although those protections vary by state and retirement system.
If investments perform poorly, the immediate consequences may fall on the pension system’s finances rather than directly on current retirees.
A fund may need higher contributions from government employers and workers. Governments may face pressure to allocate more tax revenue to retirement obligations. Some systems could adjust benefits for future employees where legally permitted.
The consequences depend on the fund’s financial condition, legal obligations, contribution policies, and investment strategy.
This is different from a 401(k), where an individual’s retirement account generally rises or falls with the investments selected.
A worker whose 401(k) is heavily concentrated in technology stocks could experience a direct decline in account value during a market downturn.
Understanding which type of retirement plan you have is therefore essential.
The Hidden Risk of Private Investments
Publicly traded technology stocks are only part of the story.
Pension funds have increasingly invested in private equity, private credit, infrastructure, and real estate. Some of those investments now help finance AI data centers, semiconductor facilities, and related infrastructure.
These assets can be more difficult to evaluate than publicly traded shares.
A company’s stock price changes continuously during market hours. Private investments may instead be valued periodically using financial models and assumptions.
That can make their reported values less responsive to changing market conditions.
Equable Institute’s 2026 State of Pensions report found that 27.1 percent of public pension assets were valued using methods other than direct market pricing.
This does not mean those investments are necessarily unsafe or improperly valued. It means that their reported prices may involve additional uncertainty.
If the economic assumptions supporting AI infrastructure change, pension managers may discover that some assets are worth less than previously estimated.
A Risk That Goes Beyond Wall Street
Artificial intelligence could influence retirement systems in another, less obvious way: through the labor market itself.
Public pensions depend partly on contributions linked to government employment and payrolls.
If AI changes the number of public employees, the kinds of jobs governments need, or the pace of wage growth, it could also affect the contributions flowing into retirement systems.
At the same time, AI might help governments improve productivity and reduce administrative costs, potentially strengthening public finances.
These possibilities are difficult to predict.
The important point is that AI may influence pensions through several channels simultaneously: investment performance, government revenue, employment, and the broader economy.
A technology that increases productivity could ultimately strengthen retirement systems. But the transition could also create financial pressures, particularly if investment expectations run ahead of economic reality.
What Retirees and Workers Should Check
Most individuals cannot decide how a large public pension fund invests its money. But they can ask informed questions about how their retirement benefits are financed and protected.
1. Find your pension fund’s annual financial report. Public retirement systems generally publish information about investments, financial obligations, and funding levels. Look for the annual comprehensive financial report or investment report.
2. Examine the funded ratio. This compares a pension system’s assets with the estimated value of benefits it has promised. A lower ratio can indicate greater financial pressure, although it does not by itself predict whether benefits will be paid.
3. Look beyond individual technology stocks. Exposure to AI may also exist through index funds, private equity, data centers, infrastructure investments, and corporate debt.
4. Understand your benefit protections. Ask your retirement administrator how your pension is calculated, what legal protections apply, and whether your benefits depend directly on market performance.
5. Review your other retirement savings. Even workers with traditional pensions may also hold 401(k)s, 403(b)s, IRAs, or other investments. Those accounts may have their own exposure to technology stocks.
The objective is not to panic or abandon investments associated with artificial intelligence. It is to understand where the risks are and how they are distributed.
The Bigger Question: Who Bears the Risk?
The AI boom has created enormous expectations for economic growth.
If those expectations are fulfilled, pension funds could benefit from years of productivity gains, business expansion, and rising corporate earnings.
But if valuations prove excessive, the losses will not be confined to technology entrepreneurs or wealthy investors.
They could also affect the public institutions responsible for financing the retirement of teachers, firefighters, police officers, and millions of other workers.
That creates a difficult responsibility for pension managers.
Avoiding AI-related companies altogether could mean missing important investment opportunities. Concentrating too heavily in them could expose retirement systems to a downturn in one increasingly interconnected sector.
The answer is not necessarily to choose between enthusiasm and fear. It is to evaluate investments according to realistic expectations, financial transparency, and the long-term obligations owed to retirees.
Artificial intelligence may indeed become one of the most transformative technologies of the century.
But even the most promising technological revolution cannot guarantee that every investment made in its name will succeed.
For millions of Americans, the real question is not whether AI will change the future. It is whether their retirement will remain secure if Wall Street’s expectations about that future turn out to be wrong.
