“AI Took Our Jobs” Was Not Accurate
In 2026, large-scale layoffs are occurring one after another in the IT industry.
Meta (approximately 8,000 people), Microsoft (early retirement incentives), Amazon (approximately 30,000 people), Oracle (up to 30,000 people)…
Every time we see such reports, many people probably think this:
“Is AI really taking human jobs after all?”
However, at least for Meta, CEO Mark Zuckerberg himself has clearly denied this view.
So what is the real reason? And is this specific to Meta, or is it an industry-wide story?
This article organizes the facts based on primary sources.
What Zuckerberg Said Internally
In May 2026, Meta announced the elimination of approximately 8,000 people, about 10% of its workforce.
Afterward, Zuckerberg directly explained to employees at an internal company meeting (town hall).
The core of his statement was this:
“Our cost structure is basically two things: computing infrastructure and personnel costs. If we increase investment in the AI field, less capital will be available for the other. So we have no choice but to shrink the size of the company.”
Even more important is this statement:
“It’s not that we’re reducing headcount because everyone in the company used AI tools to become more efficient.”
In other words, Zuckerberg is denying “layoffs as a result of AI efficiency” himself and explaining it as “layoffs as capital reallocation for AI investment.”
Why This Explanation Matters
“AI efficiency → headcount reduction” and “securing AI investment funding → headcount reduction” may look similar, but the logical structure is completely different.
In the case of layoffs due to AI efficiency
AI is introduced → human work is automated → surplus personnel emerges → reduction
In this case, layoffs occur as a “result.” People become surplus precisely because labor productivity has increased.
In the case of layoffs to secure AI investment funding
Decision to make massive investment in AI (data centers, GPUs, infrastructure) → insufficient funds → cut personnel costs
In this case, layoffs function as a “cause.” At a stage where AI is not yet generating revenue, employees are sacrificed for a bet on the future.
Zuckerberg’s explanation is the latter—that is, “headcount reduction to raise investment funds internally.”
This is neither a “benefit of the AI revolution” nor an “inevitability of technological progress,” but merely capital reallocation as a management decision.
Not Just Meta—What’s Happening Simultaneously Across the Industry
This phenomenon is not unique to Meta.
In the first quarter of 2026 alone, over 92,000 people have lost their jobs across the IT industry (according to multiple aggregations such as Layoffs.fyi).
Organizing the trends of major companies:
| Company | Reduction Scale (reported) | Official Explanation |
|---|---|---|
| Meta | Approximately 8,000 people (about 10%) | Cost reallocation accompanying AI investment expansion |
| Amazon | Cumulative approximately 30,000 people | Efficiency, AI investment expansion |
| Microsoft | Proposed early retirement to about 7% of US employees | Efficiency, cost management |
| Oracle | Thousands to up to 30,000 people (range varies by report) | Organizational restructuring, data center investment |
| Snap | Approximately 1,000 people | Reduction of repetitive tasks through AI |
However, each company’s reasons for reduction are not uniform. There are cases like Meta that explicitly state “securing funds for AI investment,” and cases like Snap that cite “business efficiency through AI” as the main cause. What’s common is the fact that “AI investment expansion and organizational restructuring are proceeding simultaneously.”
Regarding Meta, the 2026 capital expenditure (CAPEX) outlook is $125-145 billion, nearly double from the previous year’s approximately $72 billion. Meta has not denied the possibility of additional reductions.
Across the industry, major AI investment companies are rapidly expanding capital investment, and its scale is said to be unusually fast even compared to past industrial transition periods. However, comparative data sufficient to confidently say “unprecedented” is currently limited, so reservation is needed on that point.
Are “AI Efficiency Layoffs” and “AI Investment Layoffs” Different Things?
Of course, the two are not completely unrelated.
Zuckerberg himself stated that “through AI, what used to take dozens of people several months can now be done by 1-2 people in a week.”
In other words, even if it’s currently “layoffs to secure AI investment funding,” there is a high possibility it will transition to “further layoffs due to AI efficiency” in the medium to long term.
Organizing this into two phases, present and future, it looks like this:
- Present (as of 2026): Cutting personnel costs to raise funds for AI investment
- Future (several years later): Further headcount reductions progress as AI replaces work
Zuckerberg’s statement that “AI efficiency is not the cause” should be understood as a statement about the present moment.
What This Means for Employees
This structure shows a very harsh reality.
Those who benefit from AI investment are shareholders and AI infrastructure companies, and those who bear the cost are existing employees.
Meta’s 2026 capital expenditure outlook is $125-145 billion.
Even if Meta’s total employee salaries (approximately 78,000 people) were reduced to zero, the majority of this AI investment budget would still remain. Personnel cost reduction is merely a measure to supplement part of the infrastructure investment.
Even so, “personnel costs are the only thing that can be cut quickly” is the reality of today’s large IT companies.
Implications for Japanese Business Professionals
This trend is not irrelevant to Japan’s IT and tech industries.
As global major IT companies advance the “AI investment + headcount reduction” set, Japanese companies are also exposed to the following pressures:
- Competitive pressure: As overseas competitors increase productivity with AI, equivalent transformation is demanded
- Hiring changes: Recruitment for entry-level and general IT roles decreases, demand for AI specialists increases
- Skill transition: The gap between “people who can master AI tools” and “those who cannot” widens
Particularly noteworthy is the risk of becoming a “cost reduction target for AI investment” before being directly replaced by AI. Even if you think “my job can’t be done by AI,” the moment your company makes a big bet on AI, you may bear that cost.
Summary
To simply explain the IT industry layoffs occurring in 2026 as “AI took human jobs” is not accurate at the present time.
More accurately organized, it comes down to three points:
- At Meta, rather than AI efficiency itself, the aspect of headcount reduction to secure funds for massive investment in AI infrastructure is strong, and Zuckerberg himself explains it that way
- Across the industry, AI investment expansion and organizational restructuring are proceeding simultaneously, and each company’s reasons for reduction are not uniform
- In the future, as AI replaces work, headcount reductions with efficiency as a direct reason may accelerate
In other words, “AI took jobs” is only partially correct at the present time. However, we cannot say with certainty that it will remain incorrect in the future.
This phenomenon shows a phase where AI is reorganizing industries first as “an investment target” rather than as “an efficiency tool.” It may be worth watching closely what happens in the next phase.