This is a hidden option to reduce staff in "a less radical way than mass layoffs." Over the past few years, Microsoft has undergone several rounds of reductions affecting thousands of employees — most recently, last summer, the company cut 9,000 jobs.

As of last June, the company employed approximately 125,000 people in the U.S. alone. The stock buybacks will affect 7% of American personnel, which amounts to ~ 8,750 employees.

A week ago, Reuters also reported on a coming wave of large-scale layoffs, starting on May 20 — and this is not the end, but only the beginning: "subsequent cuts will begin later in the second half of the year."

Ultimately, the company plans to lay off 20% or even more of its global workforce in 2026.

Meta has spent tens of billions of dollars on attempts to create a metaverse, which ultimately failed.

In early April, Meta released Muse Spark — a model from Superintelligence Labs that was created last year. More money down the drain, because besides senseless scaling, there's nothing else to talk about: the model lacks even the very ability to reason. The company only plans to introduce a “Contemplating” mode, which will allow Muse Spark to "tackle more complex problems"
“To spend more test-time reasoning without drastically increasing latency," the company plans to... "scale the number of parallel agents."

Applause for the negative genius of the developers!

According to Amodei, finance, consulting, law, and technology are most at risk. It is in these areas that entry-level work involves research, document review, data analysis, and report preparation — routine tasks that AI performs faster and better than humans.

At Davos in January 2026, Amodei stated that AI's "cognitive breadth" means it won't destroy one industry after another.

Apparently, in order to save humanity from unemployment, Anthropic suddenly began churning out models less and less capable of "cognitively broadly" seeing the general in the particular (the level of analytics and reasoning has fallen by orders of magnitude, the price has increased).

• The proportion of junior specialists at risk of layoff reaches 50% — Axios

• The reduction in hiring of university graduates at major tech companies is almost 50% compared to pre-pandemic levels — SignalFire

• The number of AI-related layoffs in the U.S. in 2025 will be almost 55,000 — CNBC

• The share of the U.S. labor market that AI can already replace is 11.7% — CNBC citing MIT

• The number of employees fearing job loss due to AI is 40%, compared to 28% in 2024 — Mercer

In October 2025, the Yale Budget Analysis Laboratory published a report (based on U.S. labor market data from 2022 to 2025) concluding that AI has not led to mass job losses at all.

According to data from the same CNBC, the proportion of workers employed in different positions has not significantly changed since the end of 2022.

Analysts at Deutsche Bank also warned in their 2026 review that companies often use AI as a cover for job cuts, although the reasons lie in entirely different factors — primarily, the greed of the corporations themselves.