AI washing: are companies actually replacing workers with AI, or just saying they are?

Something weird is happening in corporate America. Companies are laying off tens of thousands of workers and pointing at AI as the reason. But when you look closer, the math doesn't quite add up.

The consulting firm Challenger, Gray & Christmas reported that AI was cited as responsible for more than 54,000 planned layoffs in 2025, up from roughly 17,000 combined in the two years before that. Amazon alone cut 16,000 workers in January 2026, following 14,000 reductions in October 2025. Beth Galetti, Amazon's senior VP of people experience and technology, wrote in a memo that "AI is the most transformative technology we've seen since the internet" and that the company needed to be "organized more leanly."

HP's CEO Enrique Lores said on a November 2025 earnings call that the company would use AI to "improve customer satisfaction and boost productivity," which translates to cutting 6,000 people over the coming years. Dow announced plans to lay off 4,500 employees (13% of its workforce) in January 2026, citing AI and automation as part of its operational overhaul. Meta started 2026 by cutting about 1,500 from its Reality Labs division. And xAI, Elon Musk's AI company, laid off staff in February 2026, including at least two co-founders.

That's a lot of people losing jobs. But here's where it gets murky.

The "AI washing" problem

Fabian Stephany, a research lecturer at the Oxford Internet Institute, told The Guardian that companies have a strong incentive to frame layoffs as AI-driven, even when they're not. "You can say, 'We are integrating the newest technology into our business processes, so we are very much a technological frontrunner, and we have to let go of these people,'" Stephany said.

The term for this is "AI washing." It's the corporate equivalent of greenwashing: slapping an AI label on decisions that were really made for other reasons. Cost-cutting. Pandemic-era overhiring corrections. Tariff impacts. Plain old profit maximization.

Martha Gimbel, executive director of the Budget Lab at Yale University, put it bluntly to The Guardian: "ChatGPT was only released three years ago. It is not the case that a new technology develops and the workforce adjusts immediately. That is just not how it works."

What the data actually says

A January 2026 report from Forrester projects that only 6% of US jobs will be automated by 2030. That's real, but it's not the dramatic workforce replacement that the layoff headlines suggest.

JP Gownder, a Forrester VP and principal analyst, warned that many companies are getting ahead of themselves: "A lot of companies are making a big mistake because their CEO, who isn't very deep into the weeds of AI, is saying, 'Well, let's go ahead and lay off 20 to 30% of our employees and we will backfill them with AI.' If you do not have a mature, deployed-AI application ready to do the job, it could take you 18 to 24 months to replace that person with AI — if it even works."

Meanwhile, a Resume.org survey of 1,000 US hiring managers found that 55% expect layoffs in 2026, and 44% anticipate AI will be a top driver. So even if AI washing is real, the expectation that AI will cause layoffs is baked into corporate planning.

Why this matters for developers

If you're a developer reading this, you're probably in one of two camps. Either you're worried about your job, or you're the person building the AI tools that companies claim are replacing people.

Both positions are uncomfortable in their own way.

The Business Insider reporting found something worth noting: laid-off workers across Microsoft, Amazon, and Intel are actively trying to figure out whether AI actually caused their layoffs or whether it was just the excuse. Kent Ha, a digital marketing professional laid off from Intel, saw his division's work outsourced to Accenture, which Intel said would use AI to enhance marketing. The AI framing makes the layoff feel inevitable and impersonal. Whether that framing is accurate is a separate question.

For developers specifically, the picture is more nuanced. Companies aren't replacing software engineers with ChatGPT (not yet, anyway). But they are flattening management layers, consolidating teams, and expecting remaining engineers to be more productive with AI tools. The threat isn't replacement. It's compression: fewer people expected to do more, with AI filling the gaps.

What to actually do about it

The honest answer is that nobody knows exactly how this plays out. Forrester says 6% automation by 2030. Corporate executives are acting like it's 30%. The truth is probably somewhere in between, and it'll vary wildly by role and industry.

A few things seem clear, though:

Learn the tools. Not because AI will replace you, but because companies expect you to use them. Engineers who can leverage AI assistants for code review, testing, and documentation are more valuable than those who can't.

Watch the incentives. When a company says "we're laying off workers because of AI," ask whether they've actually deployed AI systems that do those workers' jobs. Often they haven't. The layoff is about cost-cutting, and AI is the story that makes the stock price go up.

Build things that are hard to automate. System design, cross-team coordination, understanding messy real-world requirements: these are the skills that current AI handles poorly. Pure code generation is getting commoditized. The judgment around what to build and why is not.

The wave of AI-attributed layoffs will probably continue through 2026. Some of them will be genuine automation. Many will be AI washing. The hard part is telling the difference.

AI washing: are companies actually replacing workers with AI, or just saying they are?

The "AI washing" problem

What the data actually says

Why this matters for developers

What to actually do about it

Sources

Tags

Article Details