AI didn't kill the entry-level job. Interest rates and a tax change did.

The story writes itself. AI arrived. Companies started laying people off. Therefore AI caused the layoffs. It is a clean narrative, it fits on a headline, and it is largely wrong.

I have been selling AI products for three years. I have been in rooms with enterprise buyers who are simultaneously cutting headcount and buying AI software. And I can tell you that the causality is more complicated than the coverage suggests — and that understanding the real story matters enormously for how you think about your own career and how you sell into this market.

What the data actually says

A Federal Reserve study covering more than one million firms found what the researchers called “precisely-estimated null effects” for AI adoption on job postings. Not a small effect. Not a mixed effect. Null. A 2025 NBER paper tracking 25,000 workers across 7,000 workplaces found zero statistically significant effect on earnings or hours worked after AI tool adoption.

These are not fringe studies. They are among the most rigorous examinations of AI's labor market impact conducted in the post-ChatGPT era. And they do not support the displacement narrative that has dominated the conversation.

Here is what else the data shows: 59% of hiring managers in a recent survey admitted they cite AI in layoff messaging because it “plays better with stakeholders than financial constraints.” AI is being used as a narrative device to explain decisions that were made for entirely different reasons.

What actually happened

Three things converged to create the hiring freeze of 2024-2026, and none of them are named GPT-4.

Interest rates. The Fed raised rates seventeen times between 2022 and 2024. The cost of capital for growth-stage companies went from effectively zero to 5-7%. Companies that had been hiring aggressively against cheap debt capital suddenly needed to show a path to profitability. The first place most companies cut when the cost of capital rises is headcount. This is not new — it happened in 2001, in 2008, and it happened again.

Section 174.This is the one almost nobody is talking about. A 2017 tax law change took effect in 2022 that required companies to amortize R&D expenses over five years rather than deducting them immediately. For technology companies that classify significant portions of engineering salary as R&D expense — which is most of them — this dramatically changed the math on hiring engineers and developers. The entry-level tech jobs that disappeared fastest are concentrated in software engineering roles, which tracks directly with where Section 174 created the most acute cost pressure.

Post-pandemic normalization. Companies over-hired in 2020 and 2021 against pandemic-driven demand that turned out to be temporary. E-commerce companies, streaming platforms, logistics companies — all hired for a world that looked different than the one that emerged. The layoffs of 2023 and 2024 were, in significant part, the correction of that over-hiring. The timing just happened to coincide with AI becoming publicly visible.

Why this matters for how you sell

If you are selling AI to enterprise buyers right now, and you understand the above, you have a significant advantage over competitors who have swallowed the displacement narrative whole.

Enterprise buyers are not cutting jobs because AI is doing their work. They are cutting jobs because of capital costs and tax law. AI is being positioned as the justification, but the actual constraint is budget — and the actual question your buyer is asking is whether your product can help them do more with fewer people because they have to, not because the technology makes it inevitable.

That is a very different sales conversation. It is a conversation about efficiency under constraint, not transformation for its own sake. The buyers who are approving AI purchases right now are doing it to solve a budget problem, not to replace a workforce.

Lead with the outcome, not the technology. And when your buyer asks “will this replace my team,” understand that what they are often really asking is “will this help me justify to my CFO why I'm not growing headcount?” Those are different questions, and they have different answers.

The macro story in 2027 looks different from today. Rates are coming down. Section 174 is under political pressure. Post-pandemic normalization is largely complete. The companies that used the constraint period to build AI-powered workflows will come out on the other side with significant productivity advantages. The ones that waited will be hiring back into a more competitive landscape.

That is the conversation worth having with your buyers right now.