LinkedIn’s labor-market data does not yet show AI as a driver of hiring declines. Speaking at the Semafor World Economy summit, Blake Lawit, chief global affairs and legal officer at Microsoft-owned LinkedIn, said the company’s dataset shows hiring has fallen by around 20% since 2022, but that LinkedIn has “not seen” the kind of job impacts that many people associate with AI.
Lawit’s comments center on LinkedIn’s “economic graph”: if AI were already reshaping work at scale, LinkedIn’s view of jobs, skills, and hiring patterns should reveal it. Instead, Lawit pointed to rising interest rates as more closely aligned with the overall decline. At the same time, he described significant skill shifts underway and suggested that AI-driven change could arrive later.
LinkedIn’s “Economic Graph” and the Hiring Decline
Lawit grounded his argument in LinkedIn’s real-time labor-market visibility. He said LinkedIn has an “economic graph” built from over a billion members, tracking “companies, jobs, skills” to provide a view of what is happening in the labor market.
Within that dataset, Lawit said LinkedIn examined whether AI is impacting jobs right now. His conclusion was direct: “we haven’t seen it,” in reference to the question of AI’s immediate effects on hiring.
He framed the hiring slowdown as a measurable baseline trend: LinkedIn data shows “a decline in hiring of around 20% since 2022.” That figure matters because it quantifies the magnitude of change that observers might connect to AI, even if Lawit argues the connection is not supported by LinkedIn’s patterns.
Where AI Impact Would Show Up First
Lawit’s analysis relies on a specific test: if AI were already reducing hiring, LinkedIn should see that reduction where AI adoption is most expected. He said LinkedIn did not see “the sort of impacts that you would expect to see” in areas “everyone is talking about AI.” He named categories including customer support, administrative, and marketing.
Lawit added another comparative point: “Yes, hiring’s down, but not down more.” In context, this suggests LinkedIn’s data shows a decline, but not an additional decline concentrated beyond the broader downturn—at least not in the specific domains where AI-driven disruption would be expected to show first.
He also examined a demographic slice of the labor market: LinkedIn’s data did not indicate that the decline in hiring of college-aged young adults getting their first jobs was “down more,” compared with people “in the middle of or later in their careers.” That comparison is relevant because first-job hiring can be sensitive to economic conditions; Lawit’s point was that the pattern LinkedIn observed does not look like an AI-specific shock targeting early-career entry.
At the same time, Lawit did not dismiss the concern entirely. He said: “Doesn’t mean it’s not going to happen in the future, but not yet.” This suggests that LinkedIn’s dataset may show lagging effects, or that the current level of AI-driven adoption has not yet translated into the hiring patterns people expect to see.
Interest Rates as the Primary Factor
In place of AI as the immediate explanation, Lawit pointed to rising interest rates as more closely tied to the decline in hiring.
From a technology-industry standpoint, the distinction matters because it changes how companies might interpret labor-market signals. If hiring changes are primarily tied to interest rates, then AI-related workforce planning may need to account for broader economic cycles rather than treating hiring declines as evidence of immediate automation.
However, Lawit’s comments were specific to LinkedIn’s observed patterns and expectations about where AI impact would show up first. He did not provide detailed methodology beyond describing the “economic graph” and the areas examined. Observers may want to monitor how LinkedIn’s analysis evolves as AI systems become more embedded in workflows and as hiring patterns shift.
Skills Are Changing Rapidly
While Lawit argued that AI is not yet showing up as a cause of hiring declines in LinkedIn’s data, he described substantial change in the skills required for jobs. He said that over the last several years, the skills needed for the average job have changed by 25%.
With AI, LinkedIn expects that figure to rise to 70% by 2030. Lawit summarized the mechanism: “even if you’re not changing jobs, your job’s changing on you.”
This indicates that the most visible near-term effects of AI may be less about headline hiring volumes and more about job content—the mix of skills required to perform roles. For employers, that can mean rethinking training, internal mobility, and how job descriptions map to actual competency needs. For workers, it can mean that staying in a role may still require adapting to new tools and expectations, even if the company is not cutting headcount.
Lawit’s “future” warning aligns with this framing. If skill requirements are projected to accelerate, then even without an immediate AI-driven hiring drop, the labor market could experience disruption through changing requirements, shifting qualification standards, and evolving expectations for what job performance entails.
What to Watch Next
Based on Lawit’s remarks, the next signals to monitor would include whether LinkedIn’s data begins to show additional hiring declines concentrated in the areas he highlighted—customer support, administrative work, and marketing—or whether early-career hiring patterns begin to diverge from other cohorts.
Equally important, the company’s expectations about skill change—moving from 25% over recent years to 70% by 2030—could serve as a benchmark for how quickly AI-driven workflow changes propagate into job requirements. If LinkedIn’s “economic graph” continues to show large skill shifts without a matching hiring downturn, it would suggest that AI’s first-order labor effects may be transformation of roles rather than immediate net reductions in hiring.
For tech observers, the takeaway is methodological as well as substantive: LinkedIn is using large-scale labor-market data—”over a billion members” and a graph of companies, jobs, and skills—to address a question that often gets discussed with limited empirical grounding. Whether future data confirms or contradicts the “not yet” conclusion will likely depend on how quickly AI adoption translates into measurable changes in hiring and job entry.
Source: TechCrunch