The 120,000+ AI-driven layoffs tracked in 2026 concentrate in specific tiers and functions. What the pattern looks like, which roles absorb the impact, and what to do about it.
Between January and September 2026, more than 120,000 jobs were cut in layoffs directly tied to AI adoption. That number comes from 18 companies tracked on the AI layoffs tracker, and it represents only the cuts where AI was named or structurally implied. The real number is higher. Across the broader tech sector, 334 companies cut 135,000 or more roles in the same period, accounting for roughly 22 percent of all US layoffs.
These are not random headcount reductions. They follow a pattern that is visible once you know where to look: specific tiers, specific functions, specific industries. Understanding the pattern is not an academic exercise. It is a practical tool for deciding what to do next.
The pattern is tier-specific, not across-the-board
A cyclical layoff spreads proportionally across the org chart. A downturn hits revenue, the company cuts 10 percent everywhere, and the pyramid retains its shape. AI-driven layoffs look different. They concentrate inside one or two tiers while leaving the rest of the organisation intact.
In 2026, the tier absorbing the most pressure is the mid-tier production layer: the people who take a brief and produce a deliverable. Junior copywriting, mid-level financial analysis, customer service triage, marketing operations coordination, legal document review. These are roles where the daily work consists of bounded inputs producing bounded outputs inside a known format. That is exactly the work AI tools now handle reliably.
The accountability tier, the people who define the brief, review the output, own the client relationship, and carry the liability, is not being cut. In many companies it is growing. The result is a flattened pyramid: fewer production workers per senior decision-maker, with AI filling the gap between the two.
Which companies are cutting, and how much
The scale varies widely. Microsoft cut roughly 23,000 roles. UPS cut 30,000. Oracle cut 21,000. Amazon cut 16,000. Meta, IBM, PayPal, Dell, Salesforce, and Block each cut between 4,000 and 11,000. Smaller but still significant: Atlassian (1,600), Canva (1,270), Coinbase (700), Pinterest (700), Duolingo (400), Dropbox (200).
What connects these companies is not size or sector. It is the structural decision to absorb a production layer into AI tooling. UPS is a logistics company. Canva is a design platform. Coinbase is a crypto exchange. The common thread is that each had a mid-tier production function that could be compressed by AI, and each made the decision to compress it in 2026.
Five functions absorbing the most pressure
Not every function is equally exposed. Five are absorbing the majority of AI-driven cuts in 2026.
Customer service and support: chatbots, ticket routing, and automated resolution now handle the majority of Tier 1 and Tier 2 queries. The 2.9 million people employed in customer service roles in the US sit at the centre of this compression. Marketing operations: campaign assembly, performance reporting, content scheduling, and templated creative. AI handles the production; a smaller team handles the strategy. Financial analysis and corporate finance: templated modelling, variance reporting, and compliance documentation. Rule-based work that AI can execute with high accuracy. Content production: copywriting, translation, editorial production, and wire-copy journalism. Roles where output volume was the primary measure of value. Legal and compliance review: contract review, regulatory filings, due-diligence document assembly. Pattern-matching work on structured documents.
If your role sits inside one of these five functions and your daily work is primarily bounded production, the structural conditions for AI absorption are present. That does not mean a cut is imminent at your company, but it means the industry-level pattern applies to you.
How to tell the difference between AI-driven and cyclical cuts
When a layoff lands, the framing is usually vague. Efficiency. Restructuring. Alignment. The framing is designed for investors and regulators, not for the people inside the building trying to read what comes next. The short version of the diagnostic framework rests on three questions.
First, which tier got cut? If the cut concentrates at the production tier while the accountability tier stays intact, the pattern is AI absorption. Second, did a tool rollout precede the cut? AI-driven layoffs are almost always preceded by 6 to 12 months of internal AI tool deployment in the same function. Third, what does the post-cut hiring page look like? If the production tier never comes back and the senior tier keeps hiring, the strategy is permanent compression, not a temporary belt-tightening.
The roles most exposed right now
The career risk rankings score every indexed role across five structural dimensions: task routineness, decision authority, income compression, AI tool maturity, and workforce trajectory. The roles at the top of the exposure table are not a surprise once you understand the tier-specific pattern.
Customer service representatives, copywriters, translators, telemarketers, data entry clerks, brokerage clerks, and new accounts clerks all score High on structural exposure. These roles share a common profile: bounded production work, low decision authority, high task routineness, and mature AI tooling already available for their core workflow.
Medium-exposure roles are more nuanced. Roles like account executives, financial analysts, journalists, and editors have significant AI-applicable work but also carry enough judgment, relationship, or accountability to create partial insulation. For professionals in these roles, the question is not whether AI touches the work. It is whether the role can migrate upward toward the judgment layer before the production layer compresses.
What the numbers mean for the broader workforce
The 120,000 tracked cuts are the visible edge of a larger structural shift. Most AI absorption does not show up as a layoff headline. It shows up as a hiring slowdown: roles that used to be backfilled are not backfilled. Teams that used to have eight people now run with five. The work did not disappear; it was absorbed by a combination of AI tooling and the remaining team.
This silent compression is harder to track than a headline layoff, but it is arguably the larger force. The US economy adds roughly 150,000 to 200,000 jobs per month. A persistent 10 to 15 percent reduction in hiring across exposed functions is a larger displacement over a year than any single layoff event, and it is already underway in all five of the functions listed above.
What to do if your role is in the pattern
Three structural responses are available, and the right one depends on where you sit.
If you are in a high-exposure production role with no clear path to the accountability tier, the correct move is a deliberate career pivot. Not a panic quit; a structured migration into adjacent work that carries more judgment, more client contact, or more regulatory accountability. A structured positioning framework helps you decide where to migrate and how to get there.
If you are in a medium-exposure role where part of your work is automatable but part of it is not, the correct move is to deliberately shift your daily mix toward the durable part. Spend more time on judgment calls, client relationships, and original analysis. Spend less time on templated production. Make the shift visible to your manager. The people who survive a tier compression are the ones who were already operating at the tier above when the cut decision was made.
If you are in a low-exposure role, the correct move is to stay aware without overreacting. Low exposure means the structural conditions for AI absorption are not present in your work today. That can change, but it does not require immediate action.
Measure before you move
The worst response to the AI layoff pattern is an uninformed one. Moving without knowing where you actually stand leads to wasted effort, wrong pivots, and unnecessary anxiety.
Get your personal exposure score across all five dimensions. It takes about ten minutes and shows you exactly where the vulnerability sits in your role, your seniority, and your industry. The people who come through this well are not the ones who panic. They are the ones who measured first and moved deliberately.
Run your structural assessment
The AI Career Index is free to take. Get your structural snapshot in 10 minutes, including your durability score, archetype, and projection.
Take the Free Assessment