The Numbers Got More Specific — and More Alarming

Since the last reporting on this topic, the data got harder.

A working paper from the Stanford Digital Economy Lab, released in November 2025, found that workers aged 22 to 25 in AI-exposed occupations suffered a 16% relative decline in employment after generative AI spread — even after controlling for other economic factors, according to MIT Technology Review. Workers in the same occupations with more experience? No equivalent drop.

Then the Federal Reserve Bank of New York dropped this: in Q4 2025, the unemployment rate for recent college graduates hit 5.6%, while the underemployment rate — graduates working jobs that don't require a degree — climbed to 42.5%, its highest since COVID, according to MIT Technology Review.

Goldman Sachs Put a Global Number On It

As of March 18, 2026, Goldman Sachs Research estimates 300 million jobs globally are exposed to AI automation, and that in the U.S. alone, AI can automate tasks accounting for 25% of all work hours, according to Goldman Sachs.

Goldman economist Joseph Briggs — who co-leads the global economics team — says his base case assumes a 10-year adoption window, producing a 0.6 percentage point rise in unemployment. But he flagged the real risk clearly: "If it's more frontloaded, the impacts on the economy are much larger."

The overall labor market hasn't cracked yet. But Goldman confirmed what the Stanford data showed — the tech sector's employment share as a proportion of the whole economy has already fallen below its long-term trend.

AI Agents Are Already Deployed at Scale

While economists debate macro trends, AI is already doing real jobs right now.

ProCollect, a debt collection agency, is deploying AI agents — including one that calls itself "Eve" — to chase down debts at scale. According to Wired, the Kaplan Group estimates the AI debt collection industry will be worth nearly $16 billion within a decade. CareerExplorer ranks debt collection in the bottom 1% of all professions for job satisfaction. If any job was going to go first, it was this one.

Philosophers are doing better than coders. Wired confirmed that Google DeepMind now employs at least 10 in-house philosophers, Anthropic employs at least 4, and philosophy programs at Oxford and other universities are running AI ethics curricula. Henry Ajder, a philosophy postgraduate who advises the UK government and multiple startups, told Wired: "It's probably the best time to be a philosopher since Aristotle was hired as tutor to Alexander the Great."

Federal Agencies Are Classifying AI Critics as Extremists

Wired obtained more than 1,000 pages of unpublished reports from the Department of Homeland Security, the FBI, and fusion centers. The documents reveal that federal and state intelligence agencies are now classifying anti-AI sentiment as a potential extremism threat — under a category called "anti-tech violent extremism."

A New York Intelligence and Counterterrorism Bureau report obtained in that tranche states: "The chaotic atmosphere that may result from emergent AI technology in the next five years may fuel large-scale protests that devolve into civil unrest and anti-tech violent extremist activity, especially in large urban areas such as New York City."

This surveillance framework is operational. It sits on top of President Trump's National Security Presidential Memo 7, which directs the DOJ to target people holding "anti-American," "anti-Christian," and "anti-capitalism" beliefs, per Wired.

A 22-year-old who lost an entry-level software job to an AI agent, attends a protest about it, and posts about it online could now end up in a federal intelligence file.

What Mainstream Coverage Is Getting Wrong

Left-leaning outlets are running the "AI won't kill jobs" headline hard — and the macro data supports that framing for now. But they're downplaying the entry-level collapse because it complicates the reassuring narrative.

Right-leaning outlets are largely ignoring this story entirely, or treating any AI concern as Luddite panic — which aligns with the current administration's AI investment agenda.

Neither side is connecting the dots: the jobs data is deteriorating specifically for young people, the agents doing the replacing are already deployed at scale, and the government is now surveilling the people raising alarms about it.

What This Means

If you're 22 to 25 and entering the workforce in software, customer service, or any field heavy with AI exposure, the data shows you're in the hardest job market for your cohort since COVID, and AI is likely accelerating the pain.

If you're mid-career, the macro numbers say you're still fine — for now. Goldman Sachs says the 10-year clock is running.

And if you're planning to attend a protest about any of this, federal documents reviewed by Wired show federal agencies are monitoring such activity.