Hiring Is Down 20% Since 2022 — But AI Isn't the Culprit

LinkedIn has a billion-member economic graph covering companies, jobs, and skills in real time. It is among the most comprehensive labor market datasets in existence. And according to Blake Lawit, LinkedIn's chief global affairs and legal officer, that data does not show AI displacing workers — at least not yet.

Lawit made the statement at the Semafor World Economy summit this week, pushing back directly on the narrative that AI is responsible for the hiring decline the platform has observed since 2022. The actual culprit, in LinkedIn's reading: rising interest rates.

That is a significant data point in a conversation that has generated considerably more heat than light.

What LinkedIn's Data Actually Shows

Hiring on LinkedIn is down approximately 20% since 2022. That decline is real and sustained. What Lawit pushed back on is the attribution.

"We have not seen the sort of impacts that you would expect to see in areas that everyone is talking about AI — like industries, whether or not it's customer support, or administrative, or marketing," he said. If AI were driving meaningful job displacement, the signal would show up in those categories first — they are precisely the functions that AI tools are most capable of augmenting or replacing in current-generation deployments. LinkedIn's data does not show those sectors declining disproportionately.

The more parsimonious explanation is macroeconomic. Interest rate increases from 2022 onward tightened credit conditions, slowed business investment, and reduced hiring across sectors. That is a well-documented pattern in labor economics — hiring is highly sensitive to the cost of capital, and the rate environment of the past three years has been materially more restrictive than the decade that preceded it.

Lawit's summary: "Yes, hiring's down, but not down more" in the categories where AI displacement would be expected to show up first.

The Generational Angle: Entry-Level Jobs Are Not Declining Faster

One of the more specific concerns raised in public discourse about AI and employment is that entry-level positions — particularly those held by recent college graduates — are most vulnerable, because those roles tend to involve the kinds of structured, repeatable tasks that AI handles most effectively.

LinkedIn's data does not support that concern at the current moment. Lawit noted that hiring of college-aged young adults entering the workforce is not declining at a faster rate than hiring of mid-career or late-career workers. If AI were systematically eliminating entry-level work, that cohort would be the first place it would show up. It has not.

The Forward Warning: 70% Skill Change by 2030

Lawit did not suggest that the current data settles the question. His closing observation was more unsettling than the hiring decline numbers themselves.

Over the past several years, the skills required to perform the average job have changed by 25%. LinkedIn projects that figure will reach 70% by 2030 — a rate of skill transformation that has no modern precedent in the labor market.

The implication is precise: even workers whose job titles remain unchanged are doing substantially different work, requiring substantially different capabilities, within a compressed timeframe. "Even if you're not changing jobs, your job's changing on you," Lawit said.

This is the mechanism through which AI affects the labor market without necessarily reducing headcount in the short term. Jobs are not disappearing at scale — but the content of those jobs is shifting rapidly, and the workers who can adapt to those shifts are not uniformly distributed across industries, education levels, or geographies.

Why the Distinction Between Displacement and Transformation Matters

The public conversation about AI and jobs has tended to focus on displacement — the question of how many jobs AI will eliminate. LinkedIn's data suggest that framing may be premature and that the more immediate labor-market phenomenon is transformation rather than elimination.

Those are different problems requiring different responses. Displacement is a structural unemployment question — it requires policy responses around income support, retraining, and social insurance. Transformation is a question of skills and adaptation — it requires investment in continuous learning, organizational flexibility, and workforce development.

The 70% skill-change projection by 2030 is the figure that deserves the most attention from business leaders. If the skills required to perform most jobs change by 70% in four years, the organizations that have built continuous learning into their operational model will have a significant advantage over those that have not. The competitive variable is not whether your workforce uses AI — it is whether your workforce can keep pace with how AI is changing what constitutes competent performance.

What This Means for Marketing and Growth Teams

For marketing leaders specifically, Lawit's observation that marketing is one of the sectors where AI's impact is expected to show up first — and has not yet shown up in LinkedIn's hiring data — is worth sitting with.

It does not mean marketing jobs are safe from transformation. It means the transformation has not yet produced measurable displacement in the hiring data. The skill composition of marketing roles is almost certainly changing — the 25% figure applies to marketing as much as any other function — and the projected 70% by 2030 will apply there too.

The organizations building marketing teams for that future are investing in AI literacy, prompt engineering, data interpretation, and strategic judgment alongside the traditional creative and analytical capabilities that have always defined the function. The question is not whether to invest in those capabilities — it is how fast.

At Winsome Marketing, staying ahead of how AI is changing the skills and capabilities that drive growth is core to how we advise our clients. If you want to think through what that means for your team specifically, let's connect.