Microsoft announced more than 6,000 layoffs last month, the second-largest in the company’s history. As these announcements usually do these days, it came in quietly. A press statement. A memo. Since it’s almost always a Tuesday, Wednesday, or Thursday, the phrase “role impacted” will likely start to circulate through inboxes at some point between Tuesday and Thursday.
The fact that this round occurred in the same month that Microsoft’s stock began to approach all-time highs made it unique, or at the very least worthwhile. Not in the event of a collapse. Not in the midst of a financial crisis. It’s just another Wednesday. It’s the part that doesn’t quite add up, but it has practically become the norm.
Key facts & reference
| Subject | U.S. Tech Industry — Ongoing Layoff Wave (2022–2025) |
| Period covered | 2022 – present (peak in 2023; continuing through 2025) |
| Major companies involved | Microsoft, Amazon, Alphabet, Meta, Salesforce, Atlassian, Block |
| Microsoft layoffs (2025) | ~6,000 jobs cut — second largest layoff in company history |
| Peak hiring period | 2019–2022 (Amazon +106%, Meta +103% headcount growth) |
| Cited reason for cuts | AI automation, “strategic realignment,” efficiency gains |
| Most at-risk roles | Software QA, customer support, data entry, graphic design, office admin |
| Goldman Sachs estimate | ~2.5% of U.S. jobs at risk if AI deployed at maximum current capacity (2025) |
| Layoff tracker | Layoffs.fyi — real-time tech layoff database ↗ |
| Reference publication | Goldman Sachs — AI & labor market report (2025) ↗ |
The tech sector has been laying off employees since 2022 at a rate that, by all accounts, ought to have stopped by now. The initial narrative made sense: businesses like Amazon and Meta had more than doubled their workforce between 2019 and 2022 in an attempt to meet the surge in digital demand brought on by the pandemic. In retrospect, this hiring frenzy had all the discipline of a shopping spree.
The celebration ended when inflation struck and the Federal Reserve launched an aggressive tightening campaign, raising interest rates from almost zero to 5.5 percent over a fifteen-month period. It appeared that correction was required, if not inevitable. However, that adjustment was only meant to be temporary. Not a rewiring, but a reset.
Why the layoffs haven’t stopped is more difficult to explain. The cuts continue even after two years. Additionally, executives are increasingly using a novel explanation: artificial intelligence. The message, conveyed in internal memos and quarterly calls with differing degrees of polish, is that AI can now perform tasks that were previously performed by humans, and responsible management necessitates adaptation.
It’s a gripping story, in part because it tells the truth and in part because it conveniently sidesteps awkward questions about margin targets, investor pressure, and overhiring. Compared to both versions, the true story is most likely messier.
While many tasks can be automated in some way, the great majority are still primarily performed by humans, according to research released by Anthropic earlier this year. Software quality assurance, customer service, data entry, and graphic design are some job families that are more vulnerable than others, but even within those groups, actual AI displacement is still quite small.
According to Goldman Sachs’ 2025 analysis of the labor market as a whole, approximately 2.5 percent of American jobs would be seriously jeopardized if AI were implemented throughout the economy at its current maximum capacity. Not insignificant. Not, however, the massive displacement that corporate announcements frequently suggest.
It’s worth sitting with a more pessimistic interpretation of this. Approximately 75% of S&P 500 returns since ChatGPT’s launch in late 2022 have come from AI-related stocks. In addition to reducing expenses, a company that announces layoffs due to AI adoption is demonstrating to investors that it is forward-thinking, efficient, and modern.
When a workforce reduction is framed around AI, it conveys a very different message than when it is framed around declining ad revenue or a failed product cycle. It’s possible that AI is more of a cover story than the real reason behind these layoffs. Not for every business, not totally, but enough to make one wonder about the messaging’s consistency.
Examine the information from trackers such as Layoffs.FYI, and a pattern emerges rather quickly. Certain positions recur frequently in businesses of various sizes, industries, and stages. QA testers for software. tier-one customer service. roles in data entry and content tagging.
administrators of offices. coordinators of marketing. the individuals who make everything around the product work but do not actually ship the product. Even in the best of circumstances, their work is frequently invisible, making it simpler to cut when someone is arguing for efficiency.
It’s difficult to ignore the fact that some of the employees being let go were also responsible for keeping things running during the chaos of 2020 and 2021. High performers with peer reviews that said all the right things, team leads who had institutional memory nobody thought to document, engineers who built the pipelines that now run without them. Performance reviews are not the same as layoffs. These are changes to the cost structure. There is no column for loyalty or craftsmanship in the spreadsheet that powers them.
The data does indicate where the pressure is striking youngest and fastest, which is tentatively worth keeping an eye on. In the first half of 2025, unemployment increased by almost three percentage points for workers in their early twenties who entered AI-exposed fields.
Since ChatGPT’s launch, the likelihood of an unemployed person between the ages of 22 and 25 finding work in an AI-adjacent role within a month has decreased by about 14%. These figures are not disastrous. However, they have a direction. They imply that entry-level jobs, or the conventional ladder rungs, are being discreetly eliminated.
A computer science degree was almost a guarantee of six figures and several competing offers for the majority of the 2010s. With all the benefits and independence that came with competing for limited talent, tech companies spent years establishing themselves as the best places to work in the nation. A decade of almost zero interest rates underpinned that era, making talent a highly sought-after commodity and growth inexpensive.
That calculus changed as rates increased. Employees who had previously had leverage discovered that it had decreased. Additionally, layoffs, which were first presented as a correction, have gradually evolved into something more structural.
Now, it appears that the employment model is being purposefully reshaped, with a smaller workforce, AI acting as a productivity tool and justification at the same time, and the remaining employees being expected to do more with more software and fewer coworkers. Whether this is a true technological change or a financial tactic disguised in technological terminology is still up for debate. Most likely both, in ratios that differ by business and won’t be clear until later.
Not every one of the next 50,000 jobs that disappear will be apparent. There will be positions that, in the event of a resignation, are quietly unfilled. Some will be whole teams whose work is integrated into an AI workflow in a non-publicized product update. Some will be announced using the same cautious language that is almost familiar by now: operational efficiency, strategic realignment. The email sounds the same every time. At the bottom are various logos. The same gut punch.
