More than half of this year’s tech layoff announcements explicitly cite AI. Yet at the same time, more than three-quarters of technology leaders say they plan to grow permanent headcount. Both things are true — and understanding how they fit together is the key to navigating an engineering career in 2026.
If you only read the headlines, 2026 looks like the year the machines finally came for the engineers. By mid-year, analysis of layoff announcements shows that 54% of layoff events explicitly cite AI, automation or machine learning as a driving force — 164 of 302 tracked events, affecting roughly 168,770 workers. The phrase “reallocating resources towards AI” has become the default language of corporate restructuring.
But the headline number hides a more complicated, and ultimately more useful, story. The same companies cutting thousands of roles are simultaneously hiring aggressively for others. The market is not shrinking so much as violently rebalancing — away from certain kinds of work and towards others. For engineers, the practical question is not “will AI take my job?” but “which side of the rebalancing am I on, and how do I move?”
The anatomy of an AI layoff
Consider the most prominent announcements of 2026 so far. Coinbase cut around 700 employees — 14% of staff — flattening its structure to five layers below the executive team and experimenting with “one-person teams” that combine engineering, design and product in a single role; its CEO argued that engineers using AI now ship in days what previously took a team weeks. PayPal announced plans to reduce headcount by roughly 20% over two to three years — north of 4,500 roles — in a turnaround built around AI adoption. Intuit is eliminating about 3,000 jobs, around 17% of its workforce, in a restructuring centred on reallocating resources towards AI. Atlassian cut about 1,600 roles, and Meta announced a 10% reduction in April tied directly to its AI investment programme.
Two patterns stand out. First, the cuts are concentrated in work that AI tooling genuinely compresses: routine feature development, QA testing, support engineering, internal tooling and coordination-heavy middle layers. Second — and this is the part the headlines miss — several of the same companies are rehiring in different shapes. Thomson Reuters, cutting around 500 engineering roles, says it expects to refill roughly half that number within two years with senior and AI-native engineering positions. IBM, whose cumulative reductions since late 2024 exceed 15,000, is reported to be tripling its entry-level hiring for AI and hybrid-cloud roles even as it automates HR functions.
The counter-signal: demand is rising, not falling
Set against the layoff tracker, the demand data tells a strikingly different story. Robert Half’s latest research finds that 78% of technology leaders plan to increase permanent headcount in the second half of 2026 — up sharply from 61% earlier in the year — and 66% expect to bring in more contract professionals. At the same time, 65% of technology hiring managers say finding skilled talent is harder than a year ago.
The composition of demand explains the paradox. General software development postings remain far below their pre-pandemic peak — Indeed’s data has shown them down more than a third against early 2020 — while machine learning engineering openings are up 59% and AI-related postings continue to surge. Workers with multiple AI skills command a salary premium estimated at around 43% over otherwise comparable peers. The market is not paying less for engineering; it is paying differently.
| Signal | What the data shows |
| Layoff events citing AI (2026 YTD) | 54% of events; ~168,770 workers affected |
| Tech leaders planning headcount growth, H2 2026 | 78% (up from 61% earlier in the year) |
| General software development postings | Down sharply vs pre-pandemic levels |
| ML engineer openings | Up 59% |
| Salary premium for multiple AI skills | ~43% |
What is really being automated
It is worth being precise about what AI coding tools have changed. They have dramatically compressed the cost of producing code — the first draft, the boilerplate, the routine integration, the test scaffolding. What they have not compressed is the judgement around code: understanding what should be built, decomposing ambiguous problems, evaluating whether an AI-generated solution is correct, secure and maintainable, and orchestrating systems at scale.
That is why the seniority gradient in the market has steepened. Junior roles built around producing routine code are the most exposed, while senior engineers who can direct AI tooling — and take responsibility for its output — are scarcer and more valuable than ever. It also explains the rise of the “AI-native” engineer as a hiring category: someone whose default workflow assumes agentic tools, and who is measured on outcomes rather than lines of code.
A practical playbook for engineers
Make AI fluency visible, not claimed. Hiring teams increasingly screen for day-one productivity with agentic development tools. A repository, case study or interview walkthrough demonstrating how you use AI to ship faster is worth more than a skills list entry.
Move towards the judgement layer. Architecture, system design, security review, evaluation of AI outputs, and cross-functional problem framing are the tasks companies are consolidating into fewer, better-paid roles.
Consider the adjacent pivot. The shortest path to the growing side of the market is rarely a full retrain. Backend engineers move into AI engineering via RAG and integration work; QA engineers into evaluation engineering; DevOps into ML infrastructure and MLOps.
Do not ignore contract demand. With two-thirds of tech leaders planning more contract hiring, interim and project-based work is an increasingly viable route back into the market — and often the fastest way to build AI production experience.
Treat layoff language sceptically in both directions. “AI” has become a socially acceptable explanation for cuts driven partly by over-hiring, interest rates and margin pressure. Equally, a company loudly “going AI-first” may be an excellent place to build exactly the experience the market now pays for.
Key takeaways
- 54% of 2026’s layoff events cite AI, but 78% of tech leaders plan to grow permanent headcount in the second half of the year — the market is rebalancing, not collapsing.
- Cuts concentrate in routine production work; demand concentrates in AI engineering, ML infrastructure and senior judgement-heavy roles, with a ~43% premium for multi-skill AI professionals.
- The most reliable career moves are adjacent pivots — into RAG and integration work, evaluation engineering, and MLOps — backed by demonstrable AI-tooling fluency.
- Companies are rehiring in new shapes even as they cut: watch what they post, not just what they announce.