By the end of 2025, the tech layoffs that began in November 2022 had reached a scale most people now accept as normal: 123,000 people across 257 companies. What made 2025 different from 2023 and 2024 wasn’t just the volume; it was the attribution. For the first time, employers were explicitly telling people they were being laid off because of AI, with over 54,000 positions directly attributed to AI capabilities replacing human work. Simultaneously, federal government workforce reductions under DOGE directives had eliminated roughly 200,000 positions across federal agencies, with impacts cascading into defense contracting, IT services, and security clearance-dependent roles. Hardware supply chains, which had stabilized somewhat after COVID disruption, faced new chaos from tariff uncertainty and geopolitical repositioning of semiconductor manufacturing. By December, most IT professionals were facing a job market that looked fundamentally different from eighteen months prior.

The difference between 2025’s market and 2023’s lies in clarity. In 2023, the layoffs felt like overcorrection and industry consolidation; companies were admitting they’d grown too fast and were getting leaner. By 2025, it was clear that the reductions were structural and permanent. Companies weren’t planning to rehire for roles that had been eliminated. They were investing the money saved into AI tooling and automation infrastructure instead. For an IT professional, this forces a concrete reckoning: how do you position yourself in a labor market where the number of entry-level jobs is shrinking, experienced roles are becoming more specialized, and the skills that command premium salaries are not the ones that are easiest to acquire?

The 2025 Layoff Data: What It Actually Says

Start with the numbers themselves. 123,000 layoffs is real, and they’re concentrated in specific areas. Mid-market cloud infrastructure companies (think startups that built on AWS and found no moat), certain segments of ad tech, lower-tier business process outsourcing, and customer support operations took the heaviest hits. Customer support roles, in particular, were hit hard, with thousands of positions eliminated as organizations shifted to AI-powered chatbots and automated ticket triage. Basic content moderation roles, transcription, junior data entry, and data annotation roles almost entirely evaporated from the legitimate job market. These were the 54,000 roles explicitly attributed to AI.

What’s important to understand is that those 54,000 aren’t distributed randomly across IT. They’re concentrated in the lowest-skilled and most manual roles. If your career path involved moving from support into systems administration, or from data entry into analytics, that pipeline just got more congested on the entry side. This doesn’t mean those career paths don’t exist; it means the entry points are fewer and you need to be more intentional about finding them.

The federal workforce reduction was different in character. DOGE-driven reductions weren’t targeted at specific roles; they were broad-based percentage cuts across agencies. This hit not just IT roles but contracting infrastructure, clearance-dependent security roles, and the entire ecosystem of federal IT services companies. Federal contractors like Booz Allen Hamilton, Leidos, and others adjusted their growth expectations and in some cases laid off people they’d hired to support growing federal contracts. Federal IT roles that had been stable for decades suddenly became uncertain.

Who Thrived in 2025 and Why

The interesting data point from 2025 is that not all IT roles contracted. Certain specializations actually grew throughout the year and hired aggressively despite the broader market contraction. Understanding who these people are is your map for 2026.

Cybersecurity, particularly in infrastructure security, red team operations, and incident response, remained undersupplied and grew throughout 2025. Organizations didn’t cut security budgets during the broader layoffs; they often increased them because security breaches became more expensive than headcount. A CISSP-certified security engineer with five years of experience could name their salary in 2025 and get multiple offers. An AWS Solutions Architect Professional with experience in regulated industries (finance, healthcare, government) was equally in-demand. The premium for specialization was extraordinary.

Cloud infrastructure beyond entry-level roles also remained strong. Organizations needed people who could architect for multi-cloud, manage infrastructure as code at scale, and handle cloud cost optimization. Someone with deep Terraform experience, Kubernetes operational knowledge, and a track record of managing large AWS or Azure environments could find work in 2025 without breaking a sweat. The key difference was that these jobs went to people who had built that expertise, not to people trying to learn it on the job.

Data engineering, machine learning operations, and AI infrastructure roles were paradoxical: the organizations talking about replacing humans with AI were simultaneously desperate for people who could actually build and operate AI systems. A data engineer who understood how to build data pipelines that fed machine learning models, how to set up feature stores, and how to ensure data quality in production was valuable in 2025 and will be more valuable in 2026. The irony is that AI adoption requires human expertise in building the systems; the efficiency gains come later, after the infrastructure is in place.

Platform engineering and infrastructure reliability also remained in demand. Organizations realized that as they automated more, the reliability requirements for their infrastructure became more critical. A person who could build observability platforms, set up effective alerting, and design for resilience at scale was worth keeping and paying well.

The Salary Data: Where the Money Went

The data on certifications and their corresponding salaries in 2025 tells you exactly where to invest your effort. AWS Certified Solutions Architect Professional consistently commanded the highest salaries in the IT world: $221,000 average for someone with solid experience holding that cert. Kubernetes certifications (CKA, CKAD) came in at $185,000 to $195,000. CISSP (Certified Information Systems Security Professional) stayed stable at around $148,000. Terraform certifications stayed in the $140,000 range. Cloud security certifications (AWS Security Specialty, Azure Security Engineer Associate) held at $155,000 to $165,000.

Compare that to CompTIA certifications (which came in at $85,000 to $110,000) or basic cloud certifications from three years ago that didn’t correlate with meaningful salary increases anymore. The pattern is consistent: the certs that correlated with the highest salaries in 2025 were the ones that required hands-on operational knowledge, that were difficult to fake, and that directly correlated with being able to solve expensive problems for organizations. The certs that you could pass by studying a book and taking a multiple-choice exam were worth less and less.

The federal compensation story was more complex. Federal IT roles, which had been stable at $100,000 to $140,000 depending on seniority and location, faced uncertainty due to the DOGE reductions. Private sector roles for people coming out of federal service actually paid premium: someone leaving a federal GS-14 or GS-15 role could often land a private sector position at $180,000 to $220,000 because they brought operational knowledge from running large systems in production at scale, and that knowledge was valuable.

Your Concrete 2026 Playbook

Here’s what this means for you if you’re making a career move in 2026. The entry-level IT job market is tougher than it was in 2024. Support roles are automated, junior infrastructure positions are consolidated, and the “learn on the job” pathway that existed five years ago is much narrower. If you’re early in your career, you need to be strategic about where you invest your effort.

First, pick a specialization that’s hiring and defensible against automation. Cybersecurity is still the safest bet. The security skill shortage is genuine, it’s structural, and automation makes it worse rather than better because security work is inherently adversarial and requires human judgment. Infrastructure security, cloud security, or application security are all solid choices. If security doesn’t appeal to you, platform engineering and infrastructure reliability are also defensible; there’s fundamental work that requires human expertise and judgment.

Second, get credentialed in something that means something. If you’re going into security, the CISSP costs money and requires experience, but it pays for itself. If you’re going into cloud infrastructure, the AWS Solutions Architect Professional or the equivalent on Azure or GCP actually correlates with marketable skills. Don’t get me wrong: certs without skills are worthless. But certs that validate real skills are worth the investment. Budget 3K to 5K for exam costs, study materials, and hands-on lab time, and expect that investment to pay back in the first year of your career increase.

Third, if you’re mid-career, get into the habit of building things that matter. The difference between someone with a cert who can only talk about their experience and someone who can show you production infrastructure they’ve built is everything. Start a side project. Contribute to open source. Build a proof-of-concept in your organization that solves a real problem. Make it something you can talk about credibly in an interview. This is harder than studying for a cert, but it’s also more defensible against the commoditization of knowledge.

Fourth, if you’re in federal IT or contracting and your role is being disrupted, recognize that your operational knowledge is valuable in private industry, and organizations are specifically hiring for it. The people who’ve operated large systems in production at scale, who’ve managed compliance and audit processes, and who’ve worked in regulated environments have real expertise. Package that as cloud infrastructure experience, compliance knowledge, or security operations, and position yourself for the private sector. The market is hungry for people who’ve actually had to make things work, not just theoretically know how things should work.

Fifth, accept that the entry-level opportunity is in specializations or very specific technical areas, not in general IT. If you’re starting out, don’t try to be a general IT person hoping to figure out your specialization later. Pick something specific: Kubernetes operations, network security, cloud infrastructure, database administration, data engineering. Go deep on that thing. Get the cert. Build projects. Get a junior role in that specific area. That’s your entry path in 2026.

The Broader Market Shift

What’s happening in 2025 and 2026 is a maturation of the IT field. The days of broad-based, general IT skills commanding high salaries are over. The jobs that pay 150K plus are for people who have specialized expertise, who can solve expensive problems, or who have done work that’s hard to replace. The jobs at 80K to 100K are increasingly entry-level positions that are also becoming fewer, or generalist positions that are competing with AI tooling.

This forces a career choice on everyone in IT: you can try to stay general and competitive by working harder and faster, or you can specialize, build defensible expertise, and position yourself in the parts of the market that are growing. The data from 2025 strongly suggests that specialization is the winning strategy.

The Cyber Resilience Act and the broader regulatory environment are reshaping how organizations think about security hiring and capability building. If you’re considering security as a specialization, that regulatory pressure is creating both hiring demand and specific skill requirements. The 2024 Career Planning Checklist covered the mechanics of certifications and positioning that remain valid in 2026, though with the update that specialization matters more now than it did then.

2025 was a year of consolidation and efficiency for the tech industry. 2026 is going to be a year of repositioning for IT professionals. The market is clearer about what it values and what it doesn’t. If you can build specialization, demonstrate it with both certs and real projects, and position yourself where the market is hiring (security, infrastructure, data, platform engineering), you’re positioned for a career that’s resilient and well-compensated. The general IT job that your colleague might have gotten five years ago doesn’t exist anymore, but the specialized roles that pay more are there if you’re willing to build the expertise to fill them.