Engineering Jobs Are Proving Harder for AI to Displace Than Expected, New Data Suggests

For much of the last two years, the loudest prediction about artificial intelligence in the workplace has been that software development would be among the first white-collar fields to shrink. Coding assistants became faster, cheaper, and better at drafting code, while executives across the tech industry began describing AI as a force capable of replacing teams of engineers with a smaller number of highly leveraged employees. Yet new hiring data points in a different direction: software engineering may be one of the most durable job categories in tech, not the weakest.

That conclusion comes from SignalFire, a venture capital firm that tracks talent movement across the technology sector. Its latest analysis suggests that engineering has held up better than any other major function in 2025, even as the broader tech labor market remains subdued. The firm’s researchers say the evidence so far does not match the popular narrative that AI is rapidly eroding demand for developers.

The picture emerging from the data is more nuanced than a simple story of automation replacing workers. AI appears to be changing how engineers work, how productive they can be, and how companies allocate hiring budgets. But rather than eliminating the need for technical talent, the technology may be increasing the appetite for it.

Hiring data tells a different story than layoff headlines

The debate over AI and jobs has often been framed by layoffs. In May, U.S. tech layoffs reached their highest single-month total in years, and the consulting and outplacement firm Challenger, Gray & Christmas said AI was the most frequently cited reason. That headline reinforced the idea that companies are using automation to justify fewer hires and leaner teams.

SignalFire says layoff data can be misleading, in part because employment records lag reality. People often do not immediately update their professional profiles after a termination, and public layoff announcements do not always cleanly map to real workforce changes. To get a clearer read on the labor market, the firm looked at hiring instead.

Its analysis tracked the careers of millions of workers and more than 80 million companies. The result: engineering stood out as the most resilient job function in tech during 2025, despite all the fear that AI-powered coding tools would reduce demand for software talent.

“What we’re seeing on the ground is a little inconsistent with that,” said Asher Bantock, SignalFire’s head of research, referring to the common claim that AI is enabling companies to cut engineering staff dramatically.

Bantock’s view reflects a growing tension in the industry. Many executives say AI can make one engineer as productive as several. But if that were already happening at scale, the data would likely show a sharper decline in engineering hiring than in other functions. SignalFire says it does not.

Why engineering looks more resilient than other roles

According to SignalFire’s latest State of Talent Report, total hiring at large technology companies declined 25% in 2025 compared with 2019. Engineering hiring fell too, but only by 11% over the same period. In other words, engineering contracted far less than the overall tech labor market.

The report also found that engineers made up 55% of all new hires in 2025 across a group of 12 companies SignalFire labels “Tech Majors.” That group includes Alphabet, Meta, Apple, Amazon, Microsoft, Netflix, Nvidia, Tesla, Uber, Airbnb, Block and Stripe.

That share is notably higher than the 46% engineers represented in 2019, suggesting that as tech companies have become more selective, they have continued to prioritize technical talent over many other roles.

Startup hiring shows a similar pattern. At early-stage companies, engineering hiring actually rose, with startups collectively adding 7% more engineers in 2025 than they did in 2019, according to SignalFire’s data.

The message from the report is not that technology jobs are booming in general. Rather, it is that when companies must choose where to spend scarce hiring dollars, engineers remain near the top of the list.

What the numbers imply

• Tech hiring overall is still below pre-2020 levels.
• Engineering has declined less than other functions.
• Engineers now account for a larger share of new hires at major tech firms.
• Early-stage startups are still growing engineering teams.

Executives have been making bold claims about AI and labor

The SignalFire data arrives at a time when some of the most prominent figures in AI have made dramatic public warnings about the technology’s labor-market impact. Anthropic chief executive Dario Amodei said last year that AI could eliminate half of entry-level white-collar jobs and push unemployment as high as 20% within five years. That forecast helped fuel the idea that a broad white-collar contraction may be imminent.

But not all evidence from Anthropic supports that level of disruption. The company’s head of economics, Peter McCrory, told TechCrunch in March that the firm had not yet observed meaningful AI-driven changes in employment patterns.

McCrory said there was “at least no larger material difference in unemployment rates” between workers whose central tasks are heavily automated with Claude and workers in jobs less exposed to AI, such as those requiring physical dexterity and direct interaction with the world.

That distinction matters. It suggests that even where AI is being used intensively, the labor effects may be smaller, slower, or more complicated than headline-grabbing predictions imply.

Other executives have gone even further in pushing back on the idea that AI will replace engineers. Nvidia CEO Jensen Huang said in an April interview at Stanford Graduate School of Business that the notion of AI wiping out software engineering jobs was wrong. Instead, he argued, AI has made engineers busier.

Huang said that once Nvidia’s engineers began using agentic AI, they were no longer doing less work; they were doing more, because the tools created demand for additional ideas and faster iteration.

AI tools may be expanding, not shrinking, engineering work

One of the most important lessons from the current AI cycle is that productivity gains do not always reduce the need for labor. In some cases, they expand the scale of work that companies can take on. That dynamic is especially plausible in software, where lower costs and faster iteration can increase the number of features, products, experiments and infrastructure changes a company tries to pursue.

SignalFire argues that engineering is a classic example of this effect. If AI helps developers move faster, the work does not necessarily disappear. Instead, more projects can move from concept to execution, and demand rises for people who can design systems, verify outputs, integrate complex tools and manage increasingly sophisticated products.

That dynamic resembles the Jevons paradox, an economic concept that describes how efficiency gains can lead to greater overall consumption rather than less. Applied to engineering, the theory suggests that if coding becomes cheaper and faster, companies may simply do more engineering.

The result is not a world with fewer engineers, but one in which each engineer is expected to accomplish more. That shift can change team composition, junior hiring, and the mix of tasks inside a job, but it does not necessarily eliminate the job itself.

Why code generation is not the same as software development

There is a common misconception in the AI debate that writing code is the whole job. In reality, software engineering involves defining problems, making architectural decisions, handling tradeoffs, debugging systems, reviewing code, testing reliability, integrating with legacy infrastructure and working across product, design and operations teams.

AI tools can now accelerate some of those steps, particularly the drafting of boilerplate code or the generation of initial prototypes. But the broader work of shipping and maintaining software still requires human judgment, accountability and coordination.

That may help explain why engineering hiring has remained comparatively strong even as code assistants have become more capable. Companies may be using AI to reduce friction, but they have not found a way to eliminate the need for technical specialists who can own outcomes end to end.

How the labor market is being reshaped, not erased

SignalFire’s findings do not mean AI has no effect on labor. They suggest a more gradual and selective transformation. Some tasks are being automated. Some entry-level responsibilities may be changing. Some teams may be smaller than they would have been otherwise. But the strongest evidence so far does not show a wholesale collapse in engineering demand.

That is an important distinction for founders, investors and workers trying to understand the current cycle. The most visible change may not be mass job destruction, but a reallocation of work toward people who can supervise AI, validate its output and use it to move faster than competitors.

For startups, that can mean hiring fewer generalists and more technically versatile builders. For larger companies, it may mean using AI to widen the scope of what existing engineers can do while still keeping engineering as one of the largest budget priorities.

For job seekers, the implication is less straightforward. AI may pressure some categories of work more than others, and entry-level roles may face the greatest disruption. But software engineering, at least for now, appears to remain one of the safer bets inside tech.

What the latest numbers show

Why the debate over AI and jobs remains unresolved

Even with useful hiring data in hand, the broader debate is far from settled. Layoffs, hiring freezes, internal restructuring and changes in contractor use all complicate the picture. So does the fact that companies may be adopting AI at different speeds and for different purposes.

Some organizations may be using AI primarily to augment employees, while others may be testing whether it can replace tasks once handled by junior staff. In many cases, the near-term effect may be to compress roles rather than remove them entirely. One employee can oversee more output, but that does not automatically mean the company needs fewer engineers overall.

There is also a time lag problem. A technology may first show up as a productivity gain before it appears as headcount reduction. Companies often use efficiency improvements to increase product ambition, strengthen quality, or expand into new markets before they cut roles. That makes the labor-market impact slower and harder to measure in real time.

SignalFire’s work suggests that the clearest signal right now is not replacement but resilience. If AI were truly hollowing out engineering demand, the category would likely be one of the first to weaken sharply. Instead, it remains central to hiring plans across both major tech companies and startups.

What this means for the future of technical hiring

If the trend continues, engineering could become even more strategically important inside tech companies. AI may reduce the cost of some coding tasks, but it also increases the value of people who can direct that work effectively. Engineers who understand product design, systems architecture and AI-assisted development may become even more attractive hires.

The report also hints that the labor market may be separating into two stories at once. One is the decline of broad, repetitive or easily standardized work. The other is the rise of highly leveraged technical roles that sit closer to product creation and AI deployment.

That would help explain why companies still appear willing to hire engineers even while cutting elsewhere. They may see engineering not as a cost center to shrink, but as the core capability needed to turn AI from a buzzword into a business advantage.

For now, the data suggests the old assumption that coding would be the first profession to disappear may have been premature. AI has certainly changed the job, but it has not yet delivered the mass displacement many feared. Instead, engineering seems to be entering a phase in which the work is heavier, the expectations are higher and the demand remains remarkably intact.

Timeline: How the AI-and-jobs narrative has evolved

The bottom line

The most attention-grabbing claims about AI and work have often focused on what might disappear. But the latest hiring data suggests a different outcome in software: AI may be making engineers more productive without making them less necessary.

That does not rule out future disruption. It does mean that, in the current phase of the AI boom, engineering is proving more resistant than many experts predicted. For companies and workers alike, the lesson may be that the technology is changing the shape of work faster than it is eliminating it.

superintelligencenews@gmail.com