- 20
- April
In the first quarter of 2026, tech companies globally laid off nearly 80,000 workers, and the more striking statistic is that 47.9% of those positions were directly attributed to "being replaced by AI" — not the economy, not budget cuts, but a structural replacement of workers by machines.
Oracle cut 20,000-30,000; Amazon cut 16,000; Snap cut 1,000 (16% of headcount). Snap's CEO Evan Spiegel was candid in his internal memo, saying that AI enables the company to automate repetitive tasks without adding headcount. This time AI is not a secondary justification anymore — it is the primary cause, and vendors are saying so openly.
In short: Q1 2026 tech layoffs hit over 80,000 workers (~911/day), with 47.9% directly attributed to AI (not budget cuts). Oracle 20,000-30,000, Amazon 16,000, Snap 1,000 (16% of workforce), 76%+ of cuts in the U.S. — Thailand and ASEAN have a lag effect and will likely see the same impact in the next 6-18 months. Start workforce planning and AI-ready ERP preparation now.
Q1 2026 by the Numbers — Unprecedented
Layoffs in 2024 and 2025 had many justifications — "post-COVID over-hiring correction", "interest-rate pressure", "zero-based budgeting" — but Q1 2026 is clearly different. For the first time, "replaced by AI" is cited as the primary cause in nearly half of all layoff events.
| Period | Approximate Cuts | Major Companies |
|---|---|---|
| Jan 2026 | ~25,000 | Meta, Microsoft, Workday |
| Feb 2026 | ~22,000 | Atlassian (1,600), Google, Salesforce |
| Mar 2026 | ~33,000 | Oracle (major round), Amazon, Snap |
| Q1 total | ~80,000 | 47.9% cite AI as the cause |
Data from Tom's Hardware and trueup.io shows that 76%+ of the cuts are U.S.-based — U.S. labor law allows for rapid mass layoffs. As of April 19, 2026, there have been 146 layoff events totaling 99,283 workers, averaging about 911 workers per day. If this pace continues, 2026 will surpass combined 2024-2025 totals (see the broader picture at AI Layoffs 2026).
Who Cut the Most — The Top 5
The following five companies are the "headliners" of the Q1 layoff wave. Every one of them openly cited AI or "AI-driven restructuring" as the reason in their press releases or investor calls:
| Company | Cuts | % of Workforce | Public Rationale |
|---|---|---|---|
| Oracle | 20,000-30,000 | ~10-15% | Restructuring to expand AI datacenter / OCI capacity |
| Amazon | 16,000 | ~1% | AI-driven restructuring (AWS + Retail) |
| Snap | 1,000 + 300 open roles | 16% | CEO Evan Spiegel cited AI enabling automation of repetitive tasks |
| Atlassian | 1,600 | ~6% | Reallocating budget to AI projects |
| Palo Alto Networks (CyberArk M&A) | ~500 | - | Redundancy reduction post-merger + AI security automation |
Snap is the most interesting case — it is smaller than Oracle by orders of magnitude, but 16% of the workforce is a heavy proportional hit. Evan Spiegel wrote in his memo that AI tools allow smaller teams to ship product faster, reducing the need for layers of middle management and support roles. See related context at Atlassian Layoffs AI and Meta Layoffs AI.
Warning: These layoffs are structural, not cyclical. Unlike the 2024-2025 wave — driven by interest rates and post-COVID over-hiring — this one is permanent substitution. The roles hit first: customer support, QA engineers, mid-level software engineers, data entry, recruiters, technical writers. Do not expect rehiring when the market recovers — these roles are being absorbed by AI agents and LLMs, not sidelined by a tight budget (see more at AI Replacing Humans — Risks to Prepare For).
Why AI Is Replacing HR, Support, and Mid-Level Roles First
AI is not "better" at everything — it is better-suited to certain kinds of work. A role is high-risk for replacement when it has all three of these characteristics simultaneously:
- Rule-based + high-repetition — work with clear SOPs, repeatable patterns, measurable output (resume screening, ticket triage, QA test execution)
- Multi-step workflow, modest tool integration — jobs that need to pull data across 2-3 systems are now handled by LLM + agent orchestration (Claude, GPT agents reached production maturity in 2026)
- Clear cost arbitrage — an AI seat costs $20-200/month to automate work previously done by staff at $5,000-15,000/month — ROI is visible in 3-6 months
Jobs that remain "safer" tend to require local context, human judgment, and legal accountability — auditors, legal counsel, senior ERP consultants, project managers. This is exactly the thesis developed in AI vs Human: "AI being capable" does not mean AI will replace every job.
What Thai Enterprises Must Think About
Thailand and ASEAN have a "lag effect" — the same impact will arrive in 6-18 months. There are three reasons we are not yet seeing mass layoffs here:
- Thai labor law is more protective than U.S. law — terminating staff requires severance pay based on tenure (up to 400 days' wages) plus a documented reason. Mass layoffs are therefore expensive and slow, so Thai enterprises tend to use hiring freezes and natural attrition instead — shrinking the headcount without filling vacancies
- AI adoption in Thailand lags the U.S. by 12-24 months — serious enterprise AI investment only began in 2025-2026. There are plenty of tools, but few production deployments, and so the workforce impact is still muted
- Corporate culture — many Thai enterprises still view employees as "resources to take care of," not as a cost center to be optimized in the Silicon Valley style
But do not be complacent. Several Thai tech companies began freezing hiring in late 2025, and large enterprises in banking, telecom, and insurance have started piloting AI agents in place of tier-1 call center staff. There is no public layoff announcement, but vacant roles are not being refilled — the same economic result as a layoff, just quieter. This ties directly into AI + HR and Risk Management, which organizations need to plan around now. For the broader Thai sovereignty angle, see Thailand Cloud-First + GFMIS.
The Role of ERP in the AI-Driven Workforce Era
A common question: "If AI can replace this many people, is ERP still necessary?" The answer is yes — more than ever. But the role shifts from "system of record" to "infrastructure for AI agents to operate on."
Four critical roles for ERP in the AI-driven workforce era:
- HR module is still necessary — but with fewer people and different work — from "managing 1,000 employees" to "managing 500 employees + 20 AI agent seats." Payroll, leave, and performance review are still needed, but workflows must be rebuilt for a hybrid (human + AI) workforce
- Workflow automation must come before AI integration — before applying AI, processes in ERP must be standardized. Otherwise AI will automate chaos faster — not fix it
- ERP must let AI agents connect — via REST API + event-driven architecture, not vendor-specific connectors tied to a single AI provider
- Forecast workforce planning — use ERP data (attendance, productivity, project hours) to assess which roles are automation candidates, which to retain, which to reskill
Regarding Saeree ERP: honestly, our AI Assistant is still in development (training since March 2026) and not yet released to production. However, Saeree's architecture — Java + PostgreSQL + open REST APIs — already allows customers to integrate their own AI agents (Claude, GPT-based agents via API). When the Saeree AI Assistant is ready, it will build on top of the existing workflow layer. See the related data layer in HR Module.
Suitable / Not Suitable — for ERP + AI-Ready HR
Organizations are not all alike. This table helps assess whether your company is ready to invest in AI-ready ERP now:
| ✓ Invest in AI-ready ERP now | ✗ Not urgent — can wait |
|---|---|
| 300+ employees, repetitive workflows across multiple departments | Fewer than 50 employees, everyone wears multiple hats |
| HR / Admin / Support teams with highly rule-based work | Work requires heavy human judgment and local context |
| In industries where US/EU peers are already AI-adopting (finance, retail, tech services) | Industries without clear AI use cases (local services, craft goods) |
| Needs benchmarking against global peers | Local service-heavy business competing only with Thai peers |
| Legacy ERP without open APIs — migration is already needed | Current ERP is working well and has open APIs, can integrate when needed |
| CEO / CTO has AI commitment and budget ready | No clear AI strategy yet — build a roadmap before investing |
The "not urgent" column does not mean "no preparation required" — it means prepare the foundation (data quality, process standardization) first, then layer AI on top. That sequence is cheaper and safer (see related reading at China AI & Jobs and Legacy ERP Sunset).
"AI doesn't steal jobs — it gives business owners more choices between hiring humans or letting machines handle the work."
— Saeree ERP, 2026
