The biggest mistake in this debate is to describe these as straightforward “AI layoffs.” In many cases, companies are not laying off staff directly because of AI itself. Instead, they are reducing headcount while simultaneously transferring resources toward AI infrastructure, automation, data centres, and higher-margin workflows, largely due to investor pressure to increase efficiency and returns.
This indirect impact of AI is reflected in reporting: Reuters noted that Challenger, Grey & Christmas linked AI to 7% of planned U.S. layoffs in January 2026, while Goldman Sachs economists estimated that AI accounted for 5,000 to 10,000 monthly net job losses last year in the most exposed U.S. industries. At the same time, companies such as Microsoft, Oracle, Amazon, Atlassian, Autodesk, Workday, and WiseTech have specifically cited AI investment, AI-era skill shifts, or efficiency programs as factors in restructuring or job cuts. (Reuters)
The new logic of tech layoffs
This wave is about reallocation, not panic. Boards want margins protected as AI capex explodes. Microsoft and Oracle face pressure to show returns on AI spending. Amazon’s October 2025 and January 2026 cuts were tied to AI adoption and a leaner corporate structure. The pattern is jarring: profitable firms cut staff yet spend more on compute, models, and AI-enabled tooling. (Reuters)
The top 20 biggest reported layoffs by tech firms in the last 12 months
First, they are trying to fund the AI buildout without wrecking margins. AI is not cheap. It means GPUs, cloud commitments, model licensing, data engineering, security, and retraining. Microsoft’s record AI spending unsettled investors. Oracle’s restructuring is tied to rising spending on AI infrastructure. Amazon explicitly linked one of its biggest cuts to AI and to reducing bureaucracy. (Reuters)
Second, they want to change the skill mix faster than retraining allows. Atlassian said AI changes the skills and roles needed. TCS said its cuts were designed to improve agility for new technologies. WiseTech said that manual coding, as the core of engineering, is over. (Reuters)
Third, they face pressure for instant operational leverage. Block framed its 4,000-job cut as proof that a smaller team with AI tools can do more. The new boardroom math: smaller payroll, more automation, better margins. (Reuters)
The first risk is cutting muscle, not fat. If you move too fast, you lose institutional knowledge, internal trust, product memory, customer empathy, and management depth. Reuters also reported that many companies are still struggling to get meaningful returns from AI, with only 15% of executives in one Forrester survey saying profit margins improved due to AI and only 5% in a BCG survey saying they saw widespread value. In other words, many firms are making labour cuts before the payoff is proven. (Reuters)
The second risk is service quality. Klarna had to scale back its AI-only vision when customers preferred human help, and Verizon restored human service for complex cases. Aggressively cutting support roles risks harming the customer experience. (Reuters)
The third risk is damage to culture and reputation. Repeated layoffs make remaining teams slower, more political, and less inventive. People stop taking long bets when they think they may not be around to see them pay off. That effect is harder to measure than severance costs, but it is real. This is an inference, but it is consistent with the wider Reuters reporting that companies are still learning where AI genuinely works and
