Artificial intelligence could make America far wealthier by 2030. But new research raises a troubling possibility: the next generation may find the first rung of its career ladder shrinking just as the economy becomes more productive.

For 35 straight months, employment among America’s youngest workers in highly AI-exposed occupations has contracted year over year. ADP Research reported on September 23 that employment among 22- to 25-year-olds in the occupations most exposed to artificial intelligence fell 4.4% in August from a year earlier. Their peers in the least-exposed jobs lost ground too, but at less than half that pace.

Two weeks earlier, Anthropic, the company behind Claude, published a model showing how AI could make the nation dramatically richer by the end of this decade. Set those findings side by side and a hard question emerges: What good is a wealthier America if the next generation cannot get in the door?

Parents who spent years preparing their children for professional careers should take note. So should employers. The damage shows up less in layoffs than in the young people businesses never hire.

In August, Stanford economists Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen updated their study of payroll records covering millions of American workers. Employment among workers ages 22 to 25 in highly AI-exposed occupations now stands about 19% below where it would be had it kept pace with less-exposed peers of the same age. Experienced workers showed no comparable employment gap. The losses concentrate in occupations where AI automates human tasks rather than assisting the people who perform them, and the adjustment comes mainly through reduced hiring of young workers rather than increased separations.

The authors are careful. They describe these as patterns, not proof that AI caused the decline, and the gap narrows once education is taken into account. Yet it has widened steadily, from 15% in their 2025 findings to 19% by mid-2026.

Anthropic’s three scenarios frame the stakes. Under its modest scenario, AI raises 2030 output 1.6% above an economy without AI. Its substantial scenario produces an 8.3% gain. The extreme scenario lifts output 32.4%, to $44.4 trillion.

These are conditional scenarios, not forecasts. The extreme case assumes exceptionally rapid technological progress and adoption, including self-improving AI. Even so, the distribution of the gains demands a hard look.

In its substantial scenario, knowledge workers’ wages stay essentially flat while other workers see gains, and labor’s share of national output slips from about 60% to 56.1%. In the extreme scenario, knowledge-worker wages fall more than 10%, unemployment climbs past typical recession levels, and labor’s share drops to 45.2%. Total labor income barely changes even as the economy grows by nearly a third.

In that world, America could become far wealthier while many knowledge workers fall behind and a larger share of the gains flows to the owners of capital. That is not an inevitable future. But neither is it a possibility working families can afford to ignore.

Stanford’s findings strike at the traditional career ladder. For generations, young professionals learned the trade through junior positions. Accountants learned by preparing financial statements. Programmers developed expertise by writing and debugging code. Researchers acquired judgment through the demanding work of gathering information and testing conclusions. Those assignments paid the rent, but they also built the competence required for greater responsibility.

AI can now perform many tasks traditionally assigned to junior employees. The Stanford data suggest that younger workers, who often rely on knowledge acquired through formal instruction, are particularly exposed when AI substitutes for human labor, while experienced professionals, whose judgment draws on years of practice, have not suffered the same decline. But where will tomorrow’s seasoned professionals come from if businesses eliminate the very positions through which they acquire experience?

Still, the evidence does not establish that widespread AI-driven unemployment has arrived. In a May 2026 study, Yale’s Budget Lab compared employment and wages in AI-exposed occupations with those in comparable, less-exposed occupations and found no statistically or economically significant effects so far, though it cautioned that could change quickly. Yale also acknowledged its survey data are too thin to detect effects confined to recent graduates, the group Stanford flags. The Stanford team likewise finds no widespread, economy-wide displacement.

AI can complement human work rather than replace it. Anthropic’s model anticipates rising demand for occupations that benefit from greater productivity elsewhere. Faster design and permitting for infrastructure, for example, could mean more building projects and more demand for construction workers. Automating a task does not necessarily eliminate the occupation in which that task is performed.

But economic statistics cannot capture everything at stake. For Christians, work has significance beyond the paycheck. Genesis 2:15 records that God placed Adam in the Garden of Eden “to work it and keep it.” Work existed before the Fall. It belonged to humanity’s original calling, not to sin’s penalty. The Apostle Paul told believers to work heartily, “as for the Lord and not for men” (Colossians 3:23).

Our dignity does not depend upon employment or economic productivity. It comes from being created in God’s image (Genesis 1:27). No machine can diminish that dignity. Yet meaningful work remains an important way we exercise God-given abilities, develop judgment, support our families, and serve our neighbors.

A young person unable to enter a profession loses more than immediate earnings. He may lose opportunities to develop expertise, establish independence, and assume responsibilities that ordinarily accompany adulthood. Aggregate statistics hide those costs.

Rejecting technological progress is no answer, because productivity and human opportunity need not compete. Businesses should use AI to boost output without eliminating the assignments through which younger employees learn their professions — like apprenticeships, mentoring, and supervised work. Schools must teach students to use the new tools without abandoning the foundational knowledge and judgment no machine can replace. Employers who eliminate every entry-level assignment today will find their bench of experienced talent empty tomorrow.

Artificial intelligence may help America produce more wealth than previous generations could have imagined. But a growing economy is not necessarily a flourishing society. The AI economy’s success must also be measured by whether Americans retain meaningful opportunities to develop their abilities, support their families, and contribute to their communities.

Moses prayed that God would “establish the work of our hands” (Psalm 90:17). That prayer has lost none of its force in an age of thinking machines. Americans should pray and work for an economy that still leaves the next generation room to put its own hands to work.