TL;DR

SignalFire’s 2026 State of Talent Report tracked millions of employees across 80 million companies. Engineers are the most resilient function at major tech companies — 55% of all new hires in 2025, up from 46% in 2019. But the composition is shifting hard: AI/ML engineers grew 39%, front-end roles dropped 25%, and new grad hiring collapsed by 50%. The job is surviving, but the job you trained for is morphing fast.

The Panic vs. the Payroll

Every week I see the same headline recycled: AI will replace programmers, learn to prompt or perish, software engineering is dead. Dario Amodei, Anthropic’s CEO, warned that AI could eliminate half of entry-level white-collar jobs and push unemployment to 10–20% within five years. Jensen Huang told an audience at GTC that kids shouldn’t bother learning to code.

And then you look at what companies are actually doing with their money.

SignalFire, a VC firm that built its own data platform tracking 80 million companies and millions of individual career trajectories, just published its 2026 State of Talent Report. The picture it paints contradicts most of the doom headlines. Engineers aren’t getting replaced. They’re getting hired faster, leaving less often, and taking a larger share of every headcount budget than they did before the AI boom started.

I’ve watched this play out from Limassol, where the Cyprus tech scene sits at the intersection of EU regulation and Middle Eastern capital. The companies I talk to here (fintech outfits, iGaming studios, the occasional AI startup) are all hiring engineers. They’re hiring a different kind of engineer than they wanted two years ago.

What SignalFire Actually Found

The report covers the 12 companies SignalFire calls “Tech Majors”: Alphabet, Meta, Apple, Amazon, Microsoft, Netflix, Nvidia, Tesla, Uber, Airbnb, Block, and Stripe. Here are the headline numbers.

55%
of new hires are engineers (up from 46% in 2019)
−11%
engineering hiring drop vs. −25% for all roles (from 2019 baseline)
~9%
engineer attrition rate (vs. ~13% for sales and design)
+7%
more engineers hired at early-stage startups (vs. 2019)

Total hiring at the Tech Majors is down 25% from 2019 levels. But engineering hiring only dropped 11%. That gap tells you where the cuts landed: marketing, operations, support, HR, recruiting. (I tracked the broader layoff picture in the 2026 roundup.) The functions where AI tools can automate a bigger share of the work. Engineering got a smaller haircut because engineering builds and maintains those AI tools.

Early-stage startups tell the same story from the other direction. They hired 7% more engineers in 2025 than in 2019. Startups run lean by default, so when they add headcount, it’s usually the role they need most. Right now, that’s engineers.

The attrition numbers round out the picture. Engineers are turning over at roughly 9%, compared to 13% for sales and design. Companies are paying to keep them. And the engineer-to-manager ratio rose 30% over nine years (from 5.87 to 7.65). Flatter orgs, more ICs, fewer middle managers. A shift toward trusting individual contributors to operate on their own.

Winners and Losers by Specialty

The aggregate “engineers are fine” story hides a sharper realignment underneath. SignalFire broke down hiring share by engineering specialty, and the shifts since ChatGPT launched in late 2022 are dramatic.

SpecialtyChange Since 2022Direction
AI/ML Engineers+39%Growing fast
Research Engineers+28%Growing
Forward-Deployed Engineers+30%Growing
Sales Engineers+11%Growing
Front-End Engineers−25%Steepest decline

The front-end decline is the number that jumps out. A 25% drop in hiring share is the steepest of any engineering specialty. The explanation is straightforward: AI tools like Cursor, v0, and Claude Artifacts can generate, modify, and personalize UI code faster than a dedicated front-end specialist can iterate on it. Pure front-end work (slicing Figma files into React components, building form validation, responsive layouts) has become something a full-stack engineer can do with an AI assistant in a fraction of the time.

The demand migrated to people who build models, ship infrastructure, manage complex customer integrations, and sit in the room when enterprise deals close. Forward-deployed engineers grew 30% for exactly this reason: they’re the ones who help customers buy, onboard, and actually use AI products.

Meanwhile, the engineer-to-manager ratio keeps climbing. Senior IC and staff-level roles are growing as a share of hiring. Compensation structures follow the same curve. Top-of-band staff and principal packages at the Tech Majors now rival or beat director pay, reversing the management premium that defined tech careers for two decades.

The New Grad Problem

New grad hiring at the Tech Majors has dropped 50% from pre-pandemic levels. Half the entry-level pipeline vanished.

CNBC reported the same pattern: AI is absorbing the routine implementation work traditionally assigned to juniors. Companies would rather give an AI-augmented mid-level engineer a broader scope than hire two juniors and manage their ramp-up.

The market has split into a barbell, something I covered in detail in the SWE job market breakdown. Senior engineers with AI fluency are in extreme demand. Mid-level engineers who can ship independently are holding steady. Juniors without a portfolio of real projects, AI tool proficiency, and at least one shipped product are struggling to get interviews.

I wrote about this dynamic when I covered the 2026 employment data on AI replacing juniors. The SignalFire report confirms it with a different dataset. Companies want engineers, but they want engineers who can operate independently from day one. AI now handles the tasks that used to teach juniors the job.

The European Picture

SignalFire’s data skews American. The “Tech Majors” are all US-headquartered companies. But the pattern holds across the Atlantic, with some regional flavor.

About 57% of EU firms report they can’t find qualified tech staff, according to Eurostat data. Demand for AI, cloud, data, and security engineers keeps accelerating. But European companies are hiring fewer people and expecting more from each one, prioritizing mid-level and senior hires who can operate independently. The same barbell shape as the US.

Here in Cyprus (I wrote a full Cyprus developer salary breakdown earlier this year), the tech sector generates around €8.5 billion in total economic impact and supports over 62,000 full-time jobs, according to KPMG’s 2024 analysis. The 15% corporate tax rate (raised from 12.5% in January 2026), EU membership, and geographic position between Europe and the Middle East make the island attractive for fintech and iGaming. Both industries run on engineers. Junior gross pay sits around €1,000–€2,200/month depending on the stack, with fintech and ML-adjacent roles pushing €2,200–€2,800 or higher. A 50% income tax exemption for qualifying expat employees on income above €55,000 means companies can offer competitive net packages without matching London or Dubai gross salaries.

The AI-driven shift I’m seeing locally mirrors what SignalFire measured globally: companies want fewer, better engineers who can work with AI tools, and they’re willing to pay for them. The old model of hiring a large team of specialists and managing them through layers of project managers is giving way to small, capable squads where each person ships end-to-end with AI assistance.

What the People in Charge Actually Say

The public statements from tech leaders are contradictory enough to cause whiplash.

Dario Amodei (Anthropic CEO) warned AI could eliminate half of entry-level white-collar jobs and push unemployment to 10–20%. But Peter McCrory, Anthropic’s head of economics, told the same reporters that as of March 2026, they’d observed no significant AI-driven workforce effects yet. Two executives at the same company, two very different readings of the same moment.

Jensen Huang (Nvidia CEO) told audiences that “software engineers are busier than ever” with agentic AI adoption. Nvidia’s own hiring is heavily weighted toward engineers.

The disconnect makes sense once you realize that CEOs are often talking about a five-to-ten-year horizon when they predict replacement, while the hiring data reflects what’s happening right now. Right now, engineers build, maintain, integrate, and fix AI systems, plus protect the infrastructure those systems run on. That workload is growing, not shrinking.

A more accurate framing: AI will replace some of the tasks engineers do. It won’t replace engineers, because the tasks keep expanding. Every tool that automates one layer of the stack opens new work in the layers above and below it. Docker eliminated an entire class of “works on my machine” problems. It also created an entire class of container orchestration, service mesh, and cloud-native infrastructure jobs that didn’t exist before.

What Skills the Data Says to Build

If you’re reading this and wondering where to invest your time, the data points in a few clear directions.

AI skills now appear in 42% of software job descriptions, up from 8% in 2022. Engineers with two or more AI-related skills earn 43% more than peers without them. The specific tools change every six months (Copilot, Cursor, Claude Code, Codex), but the underlying skill is learning to delegate routine coding to AI and review its output critically. That skill is table stakes now.

System design and architecture are gaining a premium. AI compresses the implementation layer, so the roles that grow are the ones that decide what to build, how to structure it, and how to deliver it. Senior IC and staff roles are expanding. Architecture, distributed systems, and infrastructure experience command higher comp than they did three years ago.

Full-stack flexibility beats pure specialization. The front-end decline tracks directly to this: when AI handles UI generation well enough, a backend engineer who can also ship a frontend with AI help becomes more valuable than a frontend-only specialist. The reverse is harder. Picking up distributed systems with AI assistance is a bigger leap than picking up React.

Shipped projects beat credentials. New grad hiring is down 50%. The juniors who break through have portfolios with real deployed code, open-source contributions, and demonstrated AI fluency. A CS degree alone doesn’t cut through the noise the way it did in 2020.

Security, infrastructure, and reliability keep growing. Companies can’t automate away the need for someone who understands threat models, production incident response, and the operational reality of running services at scale. These roles resist AI substitution because they require judgment under ambiguity, which current models handle poorly.

FAQ

Will AI replace software engineers in 2026?

No. SignalFire’s data shows engineers are the most resilient hiring category at Big Tech companies, comprising 55% of new hires in 2025 with only a 9% attrition rate. The role is changing (routine implementation is increasingly AI-assisted), but companies are hiring more engineers than before.

What programming jobs are safe from AI?

Backend infrastructure, distributed systems, AI/ML engineering, security, and reliability engineering show the strongest demand growth. Forward-deployed engineering (customer-facing technical roles) grew 30% since 2022. Front-end specialist roles dropped 25%, the steepest decline of any specialty, because AI tools handle UI code generation well enough for full-stack engineers to absorb that work.

Is software engineering still a good career in 2026?

Yes, if you adapt. AI-specialized engineers average $206,000 in total base salary, and engineers with AI skills earn 43% more than peers without them. But the entry-level pipeline has narrowed. New grad hiring at the Tech Majors fell 50% from pre-pandemic levels. Breaking in requires a stronger portfolio than it did three years ago.

How will AI change software engineering?

The biggest shift is from writing code line-by-line to supervising AI agents that generate, test, and deploy code. Engineers spend less time on routine implementation and more time on architecture, code review, system design, and handling the edge cases AI stumbles on. The engineer-to-manager ratio at Big Tech rose 30%, reflecting flatter orgs where ICs operate with more autonomy.

Are software engineers in demand in 2026?

Software engineer job listings are up 30% in 2026, with over 67,000 openings tracked. But the demand profile shifted: companies want mid-to-senior engineers who can work with AI tools and ship independently. Junior roles are harder to land, and pure frontend specialization is declining. In Europe, 57% of firms report they can’t find qualified tech staff, so the headcount is there for anyone with the right skills.

Sources

Bottom Line

The data is clear enough. Engineers are getting reshuffled, not replaced. The aggregate demand is holding. The composition inside that demand is changing fast. AI/ML is up, front-end is down, new grad pipelines are half what they were, and the premium on seniority keeps growing.

Mid-career backend and infrastructure engineers are sitting in the best position. Juniors with only frontend skills and no shipped projects face a genuinely hard market. And anyone still deciding whether to learn AI tools should look at the salary data: 43% premium for engineers with AI skills, and the gap is widening.

The CEOs will keep predicting replacement. The HR departments will keep posting reqs. Somewhere between the predictions and the payroll is the truth, and right now, the payroll says engineers are fine — as long as they keep adapting to what “engineer” means in 2026.