For two years the warning has been constant: artificial intelligence would write the code, and software engineers would be the first to go. New hiring data complicates that story. According to the venture-capital firm SignalFire, which tracks employment across tens of millions of companies, engineering roles have been among the most durable jobs in technology — not the most vulnerable.
What the numbers show
In its 2026 State of Talent report, SignalFire compared hiring at a group of large technology companies it calls the "Tech Majors" — including Alphabet, Meta, Apple, Amazon, Microsoft, Netflix, Nvidia, Tesla, Uber, Airbnb, Block and Stripe — with 2019 levels. Overall hiring at those firms is down about 25%. Engineering hiring is down only about 11%, the firm reported.
The composition of hiring has shifted toward engineers, too. Software engineers made up 55% of new hires at the Tech Majors in 2025, up from 46% in 2019. Engineers are also staying put: SignalFire put their attrition rate around 9%, below the roughly 13% it found for sales and design roles. At early-stage startups the contrast is sharper still — engineering hiring was actually up about 7% versus 2019, while design and marketing fell.
"The rationale given for lots of layoffs is consistently AI," SignalFire's head of research, Asher Bantock, told TechCrunch. "What we're seeing on the ground is a little inconsistent with that."
The layoffs are real — but concentrated
The data lands against a genuinely heavy stretch of job cuts. TechCrunch's running tally of 2026 layoffs where employers cited AI includes Oracle (about 21,000 roles), Amazon (roughly 16,000 corporate jobs), Meta (around 8,000), Cisco, Intuit and Cloudflare, among others.
But the stated targets of those cuts tend not to be engineers. Cloudflare's chief executive said the great majority of its reductions fell on middle management and support functions, not engineering. Salesforce has pointed to lower customer-support volumes from its AI agents, and IBM said it had replaced a few hundred human-resources positions with AI tools. The pattern suggests that where AI is displacing workers today, it is hitting operational, administrative and support roles harder than the people who build and maintain software.
Why engineers may be holding up
Economists offer a few explanations. One is an old idea known as the Jevons paradox: when a technology makes a task cheaper, demand for that task often rises rather than falls. Applied to software, cheaper code via AI tools can expand the universe of projects worth building, sustaining demand for engineers to design, direct and review the output. Nvidia's chief executive, Jensen Huang, has argued in that vein that engineers are "busier than ever."
A second factor is that building AI systems itself requires engineers — machine-learning, platform and reliability specialists whose demand partly offsets cuts elsewhere.
The caveats
The resilience is uneven, and the report does not say AI has no effect on the labor market. Entry-level engineering roles appear more exposed than senior ones: junior developers, whose work has often meant writing straightforward code, compete most directly with AI coding tools, while senior engineers are increasingly valued for judgment and system design. Peter McCrory, an economist at Anthropic, told TechCrunch there is "at least no larger material difference in unemployment rates" for workers in AI-exposed jobs than for the broader workforce — a deliberately bounded claim, not an all-clear.
And the backdrop remains a shrunken market. A 25% drop in overall Tech Major hiring since 2019 is a real contraction; engineers' relative outperformance still leaves far fewer open roles than at the 2021–22 peak. What the data shows, for now, is narrower but notable: the mass displacement of software engineers that many predicted has not shown up in the aggregate numbers. Whether that reflects a lasting shift, a lag before deeper automation arrives, or the Jevons effect at work is, on the evidence so far, unresolved.
