Amazon is preparing fresh corporate cuts, with its People eXperience & Technology (PXT) HR group expected to be hit hardest—reports say up to 15% of PXT could go, and other orgs may see targeted reductions. Amazon hasn’t given official numbers or timing. The move lands as the company pours money into AI infrastructure and embeds automation deeper into back-office workflows.
Treat Amazon as a signal rather than an outlier. Across tech in 2025, companies are trimming roles most exposed to automation while shifting spend to AI models, data centers, and tools. By late September, trackers show tens of thousands of tech job cuts this year—even as hiring stays hot for AI engineers, applied scientists, and data-center talent.
2025 has delivered another wave of tech layoffs, with many companies explicitly citing automation and AI-driven efficiency as reasons for workforce reductions. By late September, reports based on data from Layoffs.fyi show around 90,000 jobs cut across more than 200 tech companies. At the same time, many firms are investing heavily in specialized AI roles, plus new data centers and infrastructure to power generative AI systems.
- Oracle — For example, Oracle recently laid off employees in multiple geographies — including 101 people in the Seattle area, per a state filing, plus additional announced cuts in India, the Philippines, and Canada. The job reductions come as Oracle shifts more into AI and cloud infrastructure, reallocating resources to data-center expansion and AI product development.
- Block (Square) —Another example is Block (formerly Square), which in March 2025 cut nearly 1,000 jobs, reassigning about 200 managers into non-management roles and closing around 800 open positions. The company framed it as a structural reorg to become leaner and more agile, rather than a move prompted by immediate financial concerns or replacing people with AI. That said, analysts believe automation and operational streamlining quietly factored into the decision.
- Salesforce — Another case is Salesforce, whose CEO publicly confirmed that about 4,000 customer support roles were cut this year (support staff went from ~9,000 down to ~5,000) because AI “agents” now handle large portions of customer interactions. The company claims half of its customer conversations are now managed by AI agents, freeing humans for more complex tasks, and allowing internal redeployment of roles into growth divisions.
- Fiverr — The freelance services marketplace announced plans to cut about 250 employees, roughly 30% of its workforce, as part of a restructuring to become a leaner, faster, AI-focused company. The CEO described the change as moving toward “modern, AI-focused infrastructure from the ground up,” reducing management layers while reinvesting savings into its platform.
- xAI — The AI startup headed by Elon Musk laid off around 500 data annotators, about one-third of its annotation/tutor team that helps train its chatbot Grok. Employees were notified by email, lost system access immediately, but will be paid through contract end or until November 30. Simultaneously, the company announced it will expand its specialist AI tutor roles by tenfold (in fields like STEM, finance, medicine, safety).
What these cases suggest is a broader pattern: tech firms are trimming roles that are routine or repetitive (support, operations, entry-level workflows) while scaling work in AI, model training, infrastructure engineering, and agent design. It’s not just an isolated corporate move — it’s part of a sector-wide restructuring where costs in mature functions are trimmed and resources shift toward building the AI stack.
The broader AI layoff pattern
Across tech, Amazon’s move isn’t an outlier—it’s one more data point in a broad AI-driven reshuffle. Companies are trimming roles that are easiest to automate while shifting budget to models, data centers, and tooling.
Support and back-office teams are first in line. Salesforce says AI “agents” have already shrunk its support headcount from about 9,000 to roughly 5,000, replacing repetitive triage and case work with automated workflows.
Enterprise software and fintech are following the same playbook. Oracle has run multiple layoff waves across regions as it leans harder into automation, while Block (Square) cut roughly 900–1,000 roles in a push to simplify, move faster, and do more with less.
Customer service and finance are feeling it too. In India, chatbots and voice bots are replacing large swaths of call-center work, and on Wall Street, AI rollouts are now baked into headcount plans, with firms like Goldman signaling targeted reductions as workflows get rebuilt around generative tools.
Why this is happening
At its core, this wave of layoffs reflects a shift in how value is created. The easy-to-automate parts of work—routine coordination, customer support, and administrative layers—are being replaced by AI systems that can handle structured, repetitive tasks faster and cheaper. Salesforce’s Agentforce rollout, for example, has automated roughly half of all customer interactions, letting the company operate with a much smaller human support team.
At the same time, companies are reallocating budgets toward AI infrastructure and specialist roles. Amazon’s $100 billion capex plan for 2025 is mirrored by similar spending sprees at Microsoft, Google, and Meta, all racing to expand GPU clusters and data centers. The savings from HR, operations, and mid-management cuts are being redirected to fund the physical and computational backbone of the AI era.
The message is consistent: headcount is being rebalanced, not just reduced. The priority now is fewer generalists, more engineers, data scientists, and platform experts who can design, deploy, or maintain large-scale AI systems.
How big is the shift
The numbers tell the story. Roughly 90,000 tech jobs have been cut across 200+ companies this year, according to Layoffs.fyi, but many of those same firms are simultaneously expanding their AI divisions. Salesforce, for instance, explicitly attributed a 4,000-role reduction to automation; Fiverr is cutting 30% of its staff to rebuild as an “AI-native” company; and Elon Musk’s xAI dismissed 500 data annotators while hiring specialist AI tutors in technical fields.
Globally, the trend is expected to accelerate. The World Economic Forum estimates that 41% of companies plan to reduce workforces within five years due to AI adoption. That doesn’t mean fewer jobs overall—but it does mean different ones. The shape of the modern tech workforce is shifting from service-heavy to AI-heavy, a transition that Amazon’s latest cuts—and countless smaller ones—make hard to ignore.
Conclusion
Amazon’s layoffs are part of a bigger shift sweeping through the tech industry — one where efficiency and automation take precedence over traditional headcount growth. The companies leading this transition are not cutting because they’re struggling; they’re cutting because AI lets them do more with less. Roles built around coordination, routine workflows, or volume-based operations are being automated away, while new technical roles are being created to build, manage, and scale these AI systems.
This is less a downturn than a reorganization of priorities. Money that once went into large HR or support departments is now funding GPUs, data centers, and research partnerships. Across Big Tech and smaller startups alike, the trend is unmistakable: automate the repetitive, invest in the intelligent.
The outcome will define the next chapter of the digital economy — fewer people managing processes, more people training and maintaining machines that do. Amazon’s cuts are just one signal in a much broader recalibration toward an AI-first workforce, where productivity comes less from headcount and more from computation.
