Looking for the Ladder - by Economic Innovation Group

SubscribeSign in

Looking for the Ladder<br>Is AI Impacting Entry-level Jobs?

Economic Innovation Group<br>Jan 14, 2026

38

15

Share

As part of EIG’s American Worker Project, we are delighted to present this guest post from Zanna Iscenko, AI & Economy Lead, Chief Economist's Team, Google; and Fabien Curto Millet, Chief Economist, Google. Click here for a PDF version of this post.<br>A potent narrative has taken hold in public discourse: that Artificial Intelligence (AI) is rapidly and inexorably eliminating the first rung of the career ladder for young graduates.<br>This anxiety is reflected in surveys where a majority of recent graduates find the job market challenging and believe AI is reducing the number of entry-level positions in their field. Media headlines and business reports amplify this concern, suggesting a fundamental, technology-driven restructuring of the workforce is not a distant prospect but a present reality — with entry-level, white-collar jobs as the first casualties.<br>Lending significant academic weight to this narrative is the working paper by Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen (2025), which is aptly titled “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence.”1 Its headline finding is striking: a 16 percent relative decline in employment for early-career workers (ages 22–25) in the most AI-exposed occupations since the widespread release of generative AI in November 2022.<br>While the current economic challenges of young workers in these fields are palpable and distressing, we believe that this emerging diagnosis is flawed. The most plausible explanation is that the data patterns observed are not early warnings of large-scale technological displacement, but rather the predictable consequences of a classic macroeconomic shock: the sharpest monetary policy tightening cycle in four decades.<br>Correcting the diagnosis for events to date matters, since viewing the challenges of early career workers through the narrow aperture of AI impacts could lead to overly narrow and inappropriate remedies. This does not mean, naturally, that the professional futures of young workers are safe from technological disruption going forward. Reassurance in the present should not preclude vigilance in the future. We are still in the early innings of the AI transformation, and much could happen — leading us to make recommendations for various variables that should closely be watched as part of any monitoring strategy.<br>More generally, we find the spotlight that the “Canaries” paper placed on these workers to be absolutely salutary given the profound policy questions that need addressing in areas like education and training.<br>The Trouble with Timing: Why the AI Displacement Narrative is Premature

The central weakness of the hypothesis that AI is the main driver of the recent entry-level downturn is its implausible timeline.<br>The “Canaries” paper documents a dramatic inflection point in employment for young, AI-exposed workers beginning in November 2022, immediately following the public release of ChatGPT. By June 2023, almost a half of the total observed employment decline for the occupations the paper showcases as representative of this group, software engineers and customer service workers, had already materialized.<br>This timeline suggests that within a mere six months of a consumer-facing chatbot’s launch, firms across the economy not only decided that AI could replace junior staff but also managed to implement the necessary technological infrastructure, redesigned complex workflows, ensured robust data security, and executed these staffing changes at a national scale. Such a rapid and widespread operational transformation seems implausible.<br>The reality of corporate AI adoption is a far slower and more complex process. It requires more than just employee access to a public tool. Meaningful integration that could justify replacing human labor necessitates enterprise-grade solutions that offer security guarantees, application programming interfaces (APIs) for integration into existing systems, and the development of internal expertise for effective deployment. The key tools for such an enterprise-wide shift were not available at the start of this timeline. The OpenAI API, a prerequisite for building custom applications, was only released on March 1, 2023. ChatGPT Enterprise, which sought to offer data privacy and security assurances for corporations, launched only on August 28, 2023.<br>Furthermore, the generative AI models of late 2022 and early 2023 were marked by significant reliability issues, including “hallucinations” or the fabrication of information. It is highly unlikely that risk-averse corporations would base mass staffing and hiring decisions on such nascent and unproven technology so quickly. Indeed, US Census data shows that even in Q4 2023, fewer than 10 percent of large businesses surveyed were...