Will AI Create a Second Gilded Age?

In 2026, artificial intelligence was becoming one of the most important technologies in the world.

AI systems were being used for writing, programming, research, customer service, education, design, science, entertainment, and countless other tasks.

At the same time, people were beginning to ask a much larger question.

The comparison was not necessarily a prediction that history would repeat itself.

It came from a concern that a technology capable of creating enormous amounts of economic value could also concentrate that value among the people and companies that controlled it.

Other people imagined almost the exact opposite.

They believed AI could make individuals dramatically more productive, lower the cost of goods and services, create new industries, and eventually produce an era of widespread prosperity.

In 2026, nobody knew which future would emerge.

The First Gilded Age

The term Gilded Age usually refers to the period of rapid industrial and economic growth in the United States during the late nineteenth century.

Railroads expanded. Factories grew. Industrial production increased. New industries emerged and enormous fortunes were created.

But the economic gains were distributed unevenly.

A relatively small number of industrialists accumulated extraordinary fortunes while many workers faced low wages, dangerous working conditions, long hours, and limited economic security.

The period therefore became associated with two things at once: enormous technological and economic growth, and extraordinary concentrations of wealth and power.

More than a century later, people in 2026 saw some similarities in the rise of artificial intelligence.

Why Compare AI to the Gilded Age?

AI was obviously not the same as railroads, steel, or oil.

But the economic structure surrounding AI created familiar questions.

• Who owns the technology?
• Who controls the infrastructure?
• Who receives the productivity gains?
• What happens to workers whose tasks become automated?
• Does technological progress spread wealth or concentrate it?
• How much economic power should a small number of companies have?

These questions became increasingly important as AI required enormous amounts of computing power, investment, electricity, data-center infrastructure, and specialized hardware.

The issue was therefore larger than chatbots.

AI was becoming an economic infrastructure.

What the Research Was Saying

Research available to people in 2026 did not produce a simple answer.

An International Monetary Fund working paper on AI adoption and inequality found that AI could reduce some forms of wage inequality because higher-income occupations may be more exposed to displacement. But it also found an important counter-effect: higher-income workers could benefit more from AI through increased productivity and returns on capital.

The researchers concluded that when companies choose how much AI to adopt, the resulting wealth inequality could become particularly pronounced.

In other words, AI could theoretically reduce one kind of inequality while increasing another.

That distinction mattered.

A society could have less wage inequality while simultaneously having greater inequality in who owns the assets generating economic wealth.

Other IMF research warned that AI could also increase economic differences between countries because advanced economies were generally better prepared to adopt the technology.

The Public Was Worried

The economic debate was not limited to economists. Public surveys recorded concern about AI's effects on work, but the results depend on when and where people were surveyed.

Stanford's 2026 AI Index reported a comparison from Pew surveys conducted in 2024: 73% of surveyed U.S.-based AI experts expected AI to have a positive impact on how people do their jobs, compared with 23% of U.S. adults. This is a useful baseline, but it is not a measure of public opinion in 2026.

Newer Pew polling found that 71% of U.S. adults surveyed in June 2026 expected AI to lead to fewer jobs over the next 20 years, while 5% expected more. In the same survey, 52% said they were more concerned than excited about AI's increased use in daily life. Read Pew's 2026 U.S. findings .

A Pew Research Center survey published on September 17, 2026, added a global perspective. In 34 of 37 countries surveyed, adults tended to expect AI to lead to fewer jobs rather than more. Across the countries surveyed, people more often expected AI to widen than narrow the gap between rich and poor, though many were unsure. These are expectations about AI's future effects, not measured outcomes. Read Pew's global report .

These findings show differences in expectations, not proof of which group will be right. The older expert-public comparison and the newer public surveys also measure different things and should not be treated as one continuous poll.

What Would Happen to Work?

The future of employment was at the center of the debate.

AI could make workers much more productive.

A programmer could use AI to write and debug code. A researcher could process enormous quantities of information. A designer could produce more concepts. A small business could perform work that previously required an entire department.

But productivity and employment were not the same thing.

If one worker using AI could accomplish the work previously performed by several people, companies might need fewer workers to produce the same amount of output.

That created a difficult possibility:

Whether that would be good or bad depended largely on what happened to the wealth created by the additional productivity.

The Question of Ownership

This was perhaps the deepest question underneath the entire Gilded Age comparison.

In the original Gilded Age, enormous fortunes were connected to ownership of railroads, factories, mines, oil companies, and other productive assets.

In an AI-driven economy, the important assets could look very different.

• AI models
• Computer chips
• Data centers
• Cloud computing infrastructure
• Energy infrastructure
• AI software platforms
• Data
• Intellectual property
• Companies that deploy AI at enormous scale

If these assets became extraordinarily productive, controlling them could become one of the most important sources of economic power in the world.

The AI Gilded Age vs. the Original

The comparison people in 2026 sometimes made looked roughly like this.

The two eras were obviously not identical.

The comparison was useful because it focused attention on the relationship between technological progress and ownership.

The Other Possible Future

The Gilded Age comparison represented only one possible future.

There was another vision: an age of abundance.

In this version, AI would not primarily make a small group of companies richer. Instead, it would give individuals and small organizations extraordinary productive capabilities.

• Small businesses could compete with much larger companies.
• Individuals could build products with tiny teams.
• Scientific research could accelerate.
• Education could become more personalized.
• Professional services could become cheaper.
• New industries could emerge.
• Workers could spend less time on repetitive tasks.
• Productivity could rise throughout the economy.

Economists Ajay Agrawal, Joshua Gans, and Avi Goldfarb described a similar possibility in an IMF Finance & Development article about machine intelligence.

Their argument captured an important feature of the debate: the same technology could potentially produce either broad prosperity or greater inequality depending partly on how society responded to it.

Why Concentration Mattered

AI was not simply software running on an ordinary computer.

The most advanced systems required enormous amounts of computing power, specialized chips, energy, data, engineering talent, and capital.

This created a natural concern about concentration.

If only a small number of companies could afford to build and operate the most powerful systems, those companies could potentially capture a large portion of the economic value created by AI.

Stanford's 2026 AI Index noted that frontier model production remained concentrated in the United States and China, even as open-source development expanded participation around the world.

This did not mean that a permanent AI monopoly was inevitable.

It meant that concentration was already an important part of the economic landscape people were looking at in 2026.

The Political Question

Once the discussion moved from technology to ownership, it inevitably became political.

Governments had to decide how to handle competition, taxation, worker protections, education, infrastructure, and the distribution of the economic gains from automation.

The debate was therefore no longer simply:

It increasingly became:

What People in 2026 Were Afraid Of

The fears surrounding an AI-driven Gilded Age included:

• Mass unemployment
• Extreme wealth concentration
• Declining bargaining power for workers
• AI monopolies
• Technology companies gaining excessive political influence
• A shrinking middle class
• Winner-take-all markets
• Unequal access to powerful AI systems
• Professions changing faster than workers could adapt
• Economic power becoming increasingly disconnected from human labor

What People in 2026 Hoped For

The optimistic vision was just as significant.

• Higher productivity
• Lower costs
• Shorter working hours
• Higher living standards
• New industries
• New forms of entrepreneurship
• Faster scientific discovery
• Better healthcare
• More accessible education
• More leisure time
• Greater individual economic power

In this vision, AI would become something closer to a universal tool for increasing human capability.

Instead of concentrating power, it could distribute it.

The Two Futures of 2026

Both futures could begin with exactly the same technological breakthrough.

The difference could come from what societies do with the wealth and power that follow.

AI Would Not Decide the Outcome Alone

Technology does not determine economic systems by itself.

Laws matter.

Taxes matter.

Labor markets matter.

Competition policy matters.

Education matters.

Ownership matters.

The choices governments, businesses, workers, and voters make around AI could influence whether its productivity gains become concentrated or widespread.

That was one reason the Gilded Age comparison was useful.

It moved the conversation away from simply asking what AI could do and toward asking who would benefit from what AI could do.

The Question Nobody in 2026 Could Answer

The people of 2026 had access to research, economic models, surveys, forecasts, and increasingly powerful AI systems.

But they did not have the one thing that would ultimately settle the debate.

They did not have the future.

Nobody in 2026 knew whether AI would create millions of new jobs, eliminate millions of existing ones, or do both.

Nobody knew whether AI wealth would spread throughout society or become concentrated among a relatively small group of owners.

Nobody knew whether the phrase "AI Gilded Age" would become a standard historical term or a forgotten prediction.

What a Future Reader Will Know

A person reading this article decades or centuries from now will have something that nobody in 2026 possessed.

They will know what happened.

They may know whether AI transformed employment.

They may know whether a small number of companies became extraordinarily powerful.

They may know whether AI created an era of abundance.

They may know whether inequality increased or decreased.

They may know whether governments successfully distributed the benefits of automation.

They may even know whether people in 2026 were completely wrong about the future.

But this article preserves something different.

It preserves what people thought before they knew.

One Day, This Debate Will Be History

The people of 2026 were living through the beginning of a major technological transition.

They could see AI becoming more capable.

They could see enormous amounts of money being invested into it.

They could see companies reorganizing around it.

They could see workers experimenting with it and worrying about it.

They could see researchers trying to understand its economic effects.

What they could not see was the ending.

Perhaps AI would spread economic power more widely than any previous technology.

Perhaps it would concentrate wealth and influence on an unprecedented scale.

Perhaps the future would be somewhere between the two.

In 2026, the question was still open.

One day, the answer will be history.