LinkedIn's data shows hiring is down about 20% since 2022. The drop is real. But AI isn't the obvious cause — yet.

LinkedIn's snapshot: a 20% dip

LinkedIn says its platform — which it calls an "economic graph" covering more than a billion members, companies, jobs and skills — is showing roughly a 20% fall in hiring since 2022. That's the figure Blake Lawit, chief global affairs and legal officer at LinkedIn, gave while speaking at the Semafor World Economy summit this week.

Those numbers are big enough to matter — not a blip. HR teams and policy analysts will want to know what's behind the drop.

Lawit stressed that LinkedIn's real-time view of openings, hires and skill demand hasn't flagged the pattern people expect if AI were already reshaping hiring at scale. He said LinkedIn has examined sectors often cited as vulnerable to automation — customer support, administrative roles and marketing — and hasn't seen the concentrated declines there that would point straight at generative AI as the culprit.

Interest rates, not yet AI

The explanation Lawit offered points to macro conditions rather than machine intelligence. He suggested higher interest rates are a closer match for the hiring slowdown LinkedIn is recording.

You can hear the same argument elsewhere, but it's notable coming straight from LinkedIn, which tracks these trends in real time.

Banks and employers tightened spending well before AI dominated headlines, which could explain why hiring slowed first. Hiring may have slowed because companies cut costs and paused expansion, with talk about AI following after.

Where LinkedIn looked — and didn't find AI fingerprints

LinkedIn examined categories where you might reasonably expect AI to leave a visible mark first. Those include customer support, administrative roles and parts of marketing where task automation and large language models could reduce the need for human labour. According to Blake Lawit, LinkedIn hasn't seen those sectors showing sharper falls than others.

If AI were already cutting many jobs, LinkedIn's signals would likely show it — but they don't, at least not yet.

Lawit also noted the data doesn't show a steeper hiring collapse for recent college graduates chasing first jobs compared with people in mid or late careers. Hiring is down broadly, but the decline isn't concentrated on entry-level roles in the way some feared.

Mastheads

Skills are changing — fast

That said, Lawit didn't dismiss AI's potential future impact. He reminded the audience that the mix of skills needed to do the average job has shifted about 25% in recent years. LinkedIn now expects that shift to reach about 70% by 2030 as AI and related technologies change how tasks get done.

So even if hiring numbers don't show disruption right this minute, the work itself is changing. "Even if you're not changing jobs, your job's changing on you," Lawit said.

That shifts the focus away from immediate layoffs and toward how job tasks and required skills will change over time. Employers may not be firing people en masse, but they're increasingly asking for different skills — and workers may need to retool to stay competitive.

What employers and workers might do next

When borrowing gets pricier, companies often pause hiring to save money, so that could be driving the slowdown. So the effect Lawit described — hiring down across the board, with no obvious AI hotspot — fits a classic economic slowdown pattern.

That doesn't mean firms won't later change how they deploy labour. If businesses invest in AI tools that automate parts of a job, they'll likely redraw roles, shift responsibilities and seek different combinations of human and machine skills. LinkedIn's forecast about a 70% skills shift by 2030 paints that kind of scenario.

For workers, the message is practical: the specific tasks you do today might be different in a few years even if you keep the same job title. Training, on-the-job learning and an ability to adapt to new tools could matter more than ever.

Why the caveat matters

Lawit was careful to note that what LinkedIn sees now doesn't guarantee the future. "Doesn't mean it's not going to happen in the future, but not yet," he said, acknowledging a window in which AI could become more visible in hiring data.

That nuance matters because narratives about AI and jobs can shape employer and worker behaviour. If companies expect automation to replace many roles quickly, they might slow hiring or invest differently. If workers expect to be displaced, they might rush into retraining markets or change career plans. The current data provides a counterweight to both alarm and overconfidence.

How LinkedIn's view fits into the wider debate

LinkedIn's position provides a valuable data point because the platform aggregates job postings, applications and skill mentions across a global membership. Lawit leaned on that breadth when he described the economic graph as offering a real-time market view.

That said, the platform's reach doesn't negate the possibility that localised or sector-specific shocks could already be underway but too small to move the global needle yet. Lawit didn't claim the picture is complete — only that LinkedIn's broad data hasn't found the strong AI signal some expected.

What's next for tracking AI and hiring

Companies, policymakers and labour groups will be watching several indicators: job postings by role and skill, hiring rates across cohorts, wage moves in affected occupations, and employer investment plans in automation tools. LinkedIn's data will be one of those indicators — a large one, thanks to user scale — but it won't be the only input.

The story is still unfolding, and we should keep watching hiring and skill data for clearer signs.ift quickly. New AI tools are arriving fast, and adoption patterns can accelerate once firms see clear productivity gains. For now, though, LinkedIn's day-to-day snapshot shows a hiring slowdown that fits economic tightening more than technological substitution.

So the debate over AI and jobs keeps rolling. The current evidence suggests caution about claiming a wholesale, immediate employment disruption driven by AI. But it also leaves room for significant change in how jobs get done — and who needs to learn what.

Related Articles

“Doesn’t mean it’s not going to happen in the future, but not yet,” said Blake Lawit, chief global affairs and legal officer at LinkedIn.

This article was created with AI assistance.