The Fed Just Quietly Killed Your Biggest Fear About AI and Your Paycheck

Josh Schaefer, an analyst at Barron’s, went on Fox Business’s The Bottom Line this week with a blunt message for anyone scrolling job listings while wondering if a large language model is quietly training on their replacement. His newsletter headline, he said, is essentially “AI is not taking your jobs,” and fresh research from the St. Louis Fed and the Newark Fed backs that up.

The claim is strong, and the numbers around it deserve a hard look before you either relax or ignore him.

What the Fed research actually found

Start with the labor market Schaefer is describing. The national unemployment rate sits at 4.2% as of June 2026, down from a peak of 4.5% in November 2025. Initial jobless claims came in at 215,000 for the week ending June 27, 2026, well inside the 200,000 to 250,000 range economists treat as healthy. Average hourly earnings for private-sector workers reached $37.53 in May 2026, up from $36.28 a year earlier.

Those readings describe a labor market absorbing new technology while continuing to hire. Schaefer cited a private study finding that companies investing in AI are hiring about 10% more workers over the next two years, because managers have decided a worker who can use AI to push profits higher is a worker worth keeping. The Fed’s own posture reads the same way. Policymakers have cut 75 basis points since October 2025 and have held the upper bound at 3.75% since December 11, 2025. Officials are signaling that current conditions look roughly balanced.

Schaefer’s claim is broadly correct, with one meaningful catch he flagged himself.

The catch for younger workers

The average number stops telling the truth once you split the sample by age. The St. Louis Fed research Schaefer pointed to shows that recent-graduate and undergraduate unemployment is running around 5.5%, versus the national 4.3% he cited. Recent grads usually run below the headline. Now they run above it, and that inversion is unusual looking back across the past 30 years.

Under the surface it looks like a training problem sitting inside a healthy aggregate. Entry-level postings increasingly require AI-training competency, which means the specific roles that used to teach 22-year-olds how to be useful (junior analyst, associate, coordinator) are getting reshaped before the new hire walks in the door. Work-from-home models make it harder to absorb new tools by osmosis from a senior colleague across the cube.

The mechanic is simple. AI is raising the floor on what counts as competent on day one. If you cannot clear that floor, the hiring manager picks the resume that can.

Meanwhile, real average hourly earnings adjusted for inflation sat at $11.24 in May 2026, versus $11.14 in May 2024. Purchasing power is basically flat, nominal wages are ticking up, and the jobs are there. What is uneven is who gets them.

How to stay on the right side of this

The actionable read for an individual worker is narrow and cheap.

1. Audit your own workflow. Pick the three tasks that eat the most hours in a typical week and test whether a competent AI assistant could halve any of them. An afternoon of careful testing will surface the gap faster than any think piece.
2. Build proof over vocabulary. Skip the “proficient in AI tools” line in your cover letter. Produce a one-page memo showing a real deliverable you finished faster with an AI workflow, prompts included, and it will travel through hiring pipelines on its own.
3. If you are early in your career, get in the room. Remote-first roles are efficient, and they are also where the informal tool-sharing that used to train juniors has quietly died. Hybrid or in-person exposure to how mid-career colleagues actually use these systems is currently underpriced.
4. Track the weekly jobless-claims release as your reality check. Claims moved from a September 2025 peak of 259,000 down to 215,000 by late June 2026. If that trend snaps back above 250,000 and holds, the story changes. Until it does, the aggregate data agrees with Schaefer.

The workers getting displaced right now are the ones assuming the tools will pass them by. The ones getting hired are treating fluency as table stakes.

Contact [email protected] for any questions or corrections.