AI Killed 80,000 Tech Jobs This Year... And It's Only April
Eighty thousand tech workers have lost their jobs in 2026, and the companies doing the cutting are posting record revenues. This isn't what a downturn looks like. This is something new.
The number keeps climbing and nobody seems to want to say it out loud.
Eighty thousand tech workers will lose their jobs in 2026. Eighty thousand. And it's April.
This Isn't What a Downturn Looks Like
When the hiring bubble burst in 2022 and 2023, tech companies experienced layoffs across the industry. While painful for those leaving their companies, the general economic landscape made sense.
This doesn't make that kind of sense.
Snap, the social media company behind Snapchat, recently announced that it will be laying off 1,000 employees, which would be 16% of the company’s current total workforce. The company’s CEO, Evan Spiegel, stated in a company memo that while employees will receive four months of severance, they are to continue to work from home while the company winds down. Spiegel cited one particular reason for the layoffs: AI advancements that allow the company to automate repetitive tasks. The company also cited that it will save over $500 million in the second half of the upcoming year alone.
What makes this different from previous layoffs within the tech industry is that the companies laying off workers are also some of the fastest growing companies in the sector. Snap is not dying; its closest competitor, Amazon, recently announced that its cloud AI revenue reached $15 billion this year alone. Hence, these companies are not dealing with financial distress, but have automated internal processes that no longer appear to be necessary due to the significant developments in AI technologies.
What's Actually Being Automated (It's Not What You Think)
There is a myth in the developer community that automation is the one that eats the bottom of the skill ladder first. For those who are in the middle or senior stages of their careers, this myth is comforting; it suggests that the waters are rising around the less experienced developers.
2026 reflects a messier reality.
A number of roles are being eliminated by companies that are focusing on automation tools to replace those workers who perform tasks like generating test cases, content moderation, and technical support. These tasks all involve reading a task, understanding the task, and performing a response according to some pre-established pattern - tasks that language models can perform with frightening fluency.
Companies are also eliminating middle managers for the same reasons that they are eliminating these roles: tasks like generating internal documentation, coordinating between teams, and maintaining systems can all be automated or delegated to other team members in these companies now.
The jobs that are being eliminated are not the “unskilled” jobs in the company. Instead, the jobs that are being eliminated are those that are legible to a language model. The jobs that still exist are those that require contextual judgment; someone who understands the reasons behind the behaviors of a system or who can understand the needs of a group of stakeholders when they are in disagreement about what those stakeholders require.
While such work is still, for now, performed by humans, “for now” is doing a lot of work in that sentence.
The Other Side of the Ledger
Here’s the thing: the App Store just had its biggest quarter ever.
New app releases are up 104 percent year-over-year for April 2026. On iOS alone, the growth is 89 percent. One working hypothesis, that AI coding tools are enabling people who couldn’t otherwise build software to do so, explains how a graphic designer with an idea for an app but no experience in iOS software development could build their own; how a small business owner could build an internal app without hiring a developer; or how a student with an idea and access to tools like Claude Code or Replit could finally make their app a reality with a weekend of focused effort.
The productivity gains are real. The new creation is real. The job losses are also real.
All three statements are true, and anyone who tells you only one of them is either ignorant or attempting to deceive you.
What John Chambers Said That Everyone Should Read
Chambers built Cisco during the dot-com era and watched as it inflated and burst from one of the best seats in the house. While there are similarities between the two eras, the complexity of the AI industry is even greater due to the physical infrastructure requirements.
Whereas the dot-com boom was fueled by software, the AI boom is fueled by physical infrastructure on a civilizational scale. The AI boom will cost $1.4 trillion in utilities over the next five years. Data centers are already consuming so much electricity that the power grids of several regions of the country are being strained. Furthermore, there is a shortage of transformers, again not of the AI model variety, but the electrical grid variety, slowing the deployment of this technology.
Chambers says the technology itself is real. He is referring to the fact that the infrastructure required to fuel this technology is so capital-intensive. The last time there was such a huge influx of new technology, people lost their jobs in their fields. However, new industries subsequently exploded with opportunities to employ people in these new fields.
Follow Us on X ->
View on X
Sources: TechRadar, TechCrunch, TechStartups.com, Fortune