For the past few decades, one of the most emotional economic debates in America has centered on free trade.

Supporters argued that free trade agreements would lower prices, expand markets, improve efficiency, and create new opportunities. Critics warned that the gains would not be evenly shared and that certain industries, towns, and workers would pay the price.

Both sides were partly right.

Trade helped consumers and many businesses. It also contributed to real job displacement, especially in manufacturing communities that were exposed to import competition and offshoring. The China shock, in particular, showed that trade-related job losses were not just abstract economic reshuffling. They were concentrated, painful, and slow to heal.

Now America is facing another disruption: artificial intelligence.

The question is whether AI could become the next version of the free trade shock.

The answer is yes — but not in the same way.

Free trade changed where work was done. AI changes how work is done.

Free Trade Moved Jobs. AI May Redesign Them.

 

The simplest way to compare the two is this: free trade changed where work could be done. AI changes whether certain work needs to be done by a person at all.

Free trade agreements and globalization made it easier for companies to move production, sourcing, customer service, and back-office functions across borders. The work still existed. It was often performed by people somewhere else, at a lower cost.

AI is different. It does not simply move work from one geography to another. It automates, accelerates, or compresses tasks that were previously performed by employees, contractors, agencies, call centers, analysts, writers, designers, coders, customer service reps, paralegals, salespeople, and administrative staff.

That makes the potential impact broader.

Trade-related job loss was heavily concentrated in manufacturing and import-competing industries. Manufacturing employment in the United States peaked at 19.6 million in 1979 and had fallen to 12.8 million by June 2019, according to the Bureau of Labor Statistics. Trade was not the only cause — automation, productivity gains, recessions, and business decisions also mattered — but trade pressure clearly intensified the damage in certain communities.

AI, by contrast, reaches deep into the service economy. The International Monetary Fund has estimated that about 40% of global employment is exposed to AI. In advanced economies, the exposure may be closer to 60%, with some workers benefiting from AI and others facing reduced labor demand, lower wages, or job displacement.

That does not mean 40% or 60% of jobs will vanish. Exposure is not the same as elimination. But it does mean the range of jobs touched by AI may be much wider than the range affected by traditional trade disruption.

The Similarity: The Pain Will Not Be Evenly Distributed

 

The strongest comparison between AI and free trade is not the technology itself. It is the uneven distribution of harm.

Free trade created overall economic benefits, but many of the costs were local and personal. A lower-priced product helped millions of consumers in small ways. A factory closure devastated a town in a large way.

AI could follow the same pattern.

Businesses may become more productive. Consumers may get faster service, lower prices, better tools, and more personalized experiences. Entrepreneurs may be able to do more with fewer resources. Small businesses may gain capabilities that once required large teams.

But the workers whose tasks are absorbed by AI may not experience this as progress. A company may not say, ‘We are replacing you with AI.’ Instead, it may simply hire fewer people, reduce outsourcing, consolidate roles, shrink departments, or expect one employee to do the work that previously required three.

That is exactly where the free trade comparison becomes useful. The macro story may sound positive while the individual story feels brutal.

The Difference: AI May Hit Faster and Closer to the Office

 

Free trade’s employment impact unfolded over years. Companies had to build supplier relationships, relocate production, establish overseas operations, negotiate logistics, and adjust supply chains.

AI can be adopted much faster.

A business can begin using AI tools in marketing, customer service, coding, research, proposal writing, underwriting, compliance review, lead generation, data analysis, and administrative workflows almost immediately. The implementation may not be perfect, but the barrier to experimentation is low.

That makes AI feel different from trade. It does not require a factory to close. It may show up as quiet attrition.

A company might not fire 200 people in one announcement. It may simply stop replacing people who leave. It may reduce hiring. It may push more work onto fewer employees. It may use AI to avoid adding headcount as the company grows.

This is why the early job-loss numbers may understate the long-term effect.

The OECD has noted that there is still limited evidence of broad AI-driven job losses so far, while also identifying a meaningful share of jobs as highly exposed to automation technologies, including AI. That distinction matters. The labor market may not yet show massive AI displacement, but the exposure is real.

Free Trade Hit Blue-Collar Workers First. AI May Hit White-Collar Workers First.

 

One of the most important differences is cultural and political.

Trade-related job losses were often associated with blue-collar manufacturing workers. AI disruption may be felt more strongly by white-collar workers, including people who historically felt more protected by education, credentials, and office-based careers.

That includes roles involving writing, analysis, communication, documentation, coding, customer interaction, scheduling, reporting, and repetitive decision support.

Goldman Sachs has estimated that the equivalent of 300 million full-time jobs globally could be exposed to automation by generative AI, while also noting that AI is likely to create new jobs and raise productivity.

Again, exposure does not mean elimination. But it does suggest that AI may challenge the assumption that a college degree or professional skill set automatically protects a worker from disruption.

In some ways, that makes AI more socially unsettling than trade. Trade was often framed as something that happened to factory workers. AI feels like something that can happen to almost anyone whose work is done on a computer.

The Better Historical Comparison May Be Trade Plus Automation

 

It would be too simplistic to blame manufacturing job losses only on free trade agreements. NAFTA, for example, remains heavily debated. A Congressional Research Service report concluded that NAFTA did not produce either the huge job losses feared by critics or the large gains promised by supporters, and that its net overall effect on the U.S. economy was relatively small.

The larger lesson is that job loss usually does not come from one force alone.

Manufacturing workers were hit by a combination of trade, automation, productivity gains, corporate restructuring, weak retraining systems, and regional economic dependence on single industries.

AI may create a similar combination.

It will not act alone. It will combine with cost pressure, remote work, outsourcing, software platforms, private equity efficiency mandates, inflation, wage pressure, and management’s desire to do more with less.

That is why the free trade comparison is useful — not because the mechanics are identical, but because the pattern may rhyme.

The economy may become more efficient while certain workers become less secure.

The Main Lesson: Do Not Dismiss the Disruption Because the Economy Benefits Overall

 

One of the mistakes made during the free trade era was assuming that because the economy benefited in aggregate, displaced workers would adjust smoothly.

Many did not.

Research on the China shock found that labor market adjustment was slow, with wages and labor-force participation remaining depressed in affected areas. In other words, workers did not simply move, retrain, and quickly reappear in better jobs.

That lesson should be applied to AI.

It is not enough to say, ‘AI will create new jobs.’ It probably will. Most major technologies do. But the new jobs may not appear in the same places, pay the same wages, require the same skills, or be available to the same people who lost the old jobs.

That was the painful lesson of free trade.

It could become the painful lesson of AI.

The Business View: AI Will Be Too Valuable to Ignore

 

From a business standpoint, AI adoption is not going to stop. Companies that use AI well may gain speed, efficiency, better customer service, improved decision-making, and lower operating costs.

That creates a competitive reality.

If one company can use AI to respond to customers instantly, generate proposals faster, analyze data more accurately, reduce administrative friction, and operate with fewer bottlenecks, competitors may be forced to follow.

This is another parallel with free trade. Once companies discovered lower-cost global supply chains, many competitors had little choice but to participate. The same may happen with AI-powered operations.

The issue is not whether AI will be adopted.

It will.

The issue is whether companies, workers, educators, and policymakers prepare for the transition more intelligently than they did with trade.

Conclusion: The Comparison Is Real, But AI Could Be Broader

 

AI-related job loss should not be described as simply another version of free trade job loss. That would be too narrow.

Free trade primarily exposed workers to lower-cost labor elsewhere.

AI exposes workers to software that can perform, assist, or accelerate tasks once believed to require human judgment, communication, creativity, or analysis.

The comparison is strongest in the human and economic pattern: broad gains, concentrated losses, uneven adjustment, and a tendency to underestimate the pain of workers caught in the middle.

Free trade taught us that economic progress can still leave people behind.

AI may teach the same lesson again — only this time, the disruption may not be limited to factories, imports, or industrial towns.

It may reach into the office, the call center, the agency, the professional firm, the software company, the finance department, the sales team, and the small business.

The real question is not whether AI will destroy every job. It will not.

The real question is whether AI will quietly change enough jobs, fast enough, that millions of workers feel the ground shift beneath them before the economy has figured out how to help them land somewhere better.

 

Selected Sources

 

• David Autor, David Dorn, and Gordon Hanson, “The China Syndrome: Local Labor Market Effects of Import Competition in the United States,” American Economic Review. https://www.aeaweb.org/articles?id=10.1257/aer.103.6.2121
• U.S. Bureau of Labor Statistics, “Forty years of falling manufacturing employment.” https://www.bls.gov/opub/btn/volume-9/forty-years-of-falling-manufacturing-employment.htm
• International Monetary Fund, “AI Will Transform the Global Economy. Let’s Make Sure It Benefits Humanity.” https://www.imf.org/en/blogs/articles/2024/01/14/ai-will-transform-the-global-economy-lets-make-sure-it-benefits-humanity
• OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market. https://www.oecd.org/employment-outlook/2023/
• Goldman Sachs, “Generative AI could raise global GDP by 7%.” https://www.goldmansachs.com/insights/articles/generative-ai-could-raise-global-gdp-by-7-percent
• Congressional Research Service, “NAFTA at 20: Overview and Trade Effects.” https://crsreports.congress.gov/product/pdf/R/R42965
• National Bureau of Economic Research, “The China Shock: Learning from Labor Market Adjustment to Large Changes in Trade.” https://www.nber.org/papers/w21906