On February 26, 2026, Jack Dorsey posted a message that sent shockwaves through Silicon Valley. Block, the payments company he built from a card reader and a dream, was cutting 4,000 employees. Nearly half the workforce, gone in a single day. The reason Dorsey gave was blunt: AI tools had made those jobs unnecessary. “A significantly smaller team, using the tools we’re building, can do more and do it better,” he wrote. Wall Street cheered. Block’s stock jumped 24% in after-hours trading.
But within days, a different narrative started forming. Bloomberg published an investigation questioning whether AI had anything to do with it. Sam Altman, the CEO of OpenAI and arguably the person closest to AI’s actual capabilities, told CNBC at the India AI Impact Summit: “There’s some AI washing where people are blaming AI for layoffs that they would otherwise do.” When the guy building the technology says companies are exaggerating its impact, it’s worth paying attention.
AI washing is when companies exaggerate or fabricate the role of artificial intelligence in their products, services, or workforce decisions to appear more innovative or justify cost-cutting that has nothing to do with automation.
Key Takeaways
- AI washing takes two forms: companies slapping “AI-powered” labels on products that aren’t, and companies blaming AI for layoffs driven by overhiring corrections and cost-cutting.
- The SEC imposed its first AI washing penalties in March 2024, fining investment firms Delphia ($225,000) and Global Predictions ($175,000) for fabricating AI capabilities.
- Challenger, Gray & Christmas data shows AI was cited in 12,304 U.S. job cuts in the first two months of 2026, but a National Bureau of Economic Research paper found 90% of executives say AI has had zero impact on employment at their companies.
- Block cut 4,000 employees citing AI in February 2026, while Meta reportedly considered cutting 20% of its workforce to offset AI infrastructure spending, not because AI replaced those workers.
- OpenAI CEO Sam Altman publicly confirmed that companies are using AI as cover for financially motivated layoffs, saying “almost every company that does layoffs is blaming AI, whether or not it really is about AI.”
Last updated: March 2026
What is AI washing?
AI washing borrows its name from greenwashing, the decades-old practice of companies pretending to be more environmentally friendly than they are. The mechanics are identical. Where greenwashing slaps a green leaf on a product made in a coal-powered factory, AI washing slaps “powered by AI” on software running basic if-then logic.
The term originally described a narrow problem in marketing and securities. Companies would tell investors their products used sophisticated machine learning when the underlying technology was far simpler, or sometimes nonexistent. But in early 2026, the meaning expanded. AI washing now also describes companies that cite artificial intelligence as the reason for mass layoffs when the real drivers are pandemic-era overhiring, falling revenue, or pressure from activist investors.
Both versions share a core deception: using the mystique of AI to make a business decision look smarter and more forward-thinking than it actually is.
The two types of AI washing
Product AI washing
This is the original flavor. A company markets its product as “AI-powered” when the technology under the hood is either basic automation, rule-based software, or humans doing the work behind a digital curtain.
Amazon’s Just Walk Out technology became the poster child for product AI washing. The system was marketed as AI that could track what shoppers picked up and charge them automatically. In reality, as many as 1,000 workers in India were reviewing video footage of customers and manually logging purchases. In 2022, 700 out of every 1,000 transactions required human verification. Amazon quietly retired the system in 2024.
The Rabbit R1, a handheld device marketed as a standalone AI assistant, sold out its first production run. Teardowns revealed it was essentially running ChatGPT with a few hardcoded scripts. Coca-Cola released a product called Y3000 in 2023, claiming the flavor was “co-created with artificial intelligence.” No one could explain what that meant, because it didn’t mean anything.
Employment AI washing
This is the 2026 version, and it’s bigger. Companies announce massive layoffs and tell the press it’s because AI made those roles obsolete. The layoffs are real. The AI explanation often isn’t.
Block’s February 2026 cuts are the clearest example. Dorsey said AI tools could do the work of 4,000 people. But Block’s own financial filings showed the company had grown from 5,477 employees in 2020 to over 10,000 by 2025 during a hiring spree that tracked pandemic-era fintech hype, not sustainable demand. The cuts brought headcount roughly back to where it was before the over-expansion.
Meta’s reported plans to cut up to 20% of its workforce follow the same pattern. The company framed the potential layoffs as a response to AI infrastructure costs, not as AI replacing workers. Put differently: Meta isn’t laying people off because AI does their jobs. Meta is laying people off to pay for the servers that run AI.
Fortune’s Dan Abelon captured the dynamic bluntly: Mark Zuckerberg is positioned to finish what Jack Dorsey started, a “cascade” of AI-related layoffs across the tech sector where the AI justification is often more narrative than reality.
Is AI washing illegal?
Yes, at least the product version. The SEC cracked down on AI washing starting in March 2024 with penalties against investment advisors Delphia and Global Predictions. Delphia paid $225,000 for claiming it used AI and machine learning to analyze client data when it never actually did so. Global Predictions paid $175,000 for similar fabrications.
The enforcement escalated in 2025. The SEC charged Presto Automation, the first publicly traded company hit, for claiming its restaurant ordering technology was AI-driven when it actually relied on human agents operating behind the scenes. Then came the criminal case: the SEC and Department of Justice charged the founder of Nate Inc. with raising over $42 million by claiming his shopping app used AI to process transactions when the company actually employed manual workers. The app’s supposed 90% automation rate was essentially zero.
The FTC joined the fight with Operation AI Comply, targeting companies using AI claims to sell fake review generators, bogus “AI lawyer” services, and dropshipping schemes that promised AI-powered passive income.
Employment AI washing exists in a legal gray area. No company has been charged for citing AI as a layoff reason that wasn’t entirely accurate. But securities lawyers are watching. If a company tells investors that AI-driven efficiency improvements justify workforce cuts, and those improvements don’t exist, that’s potentially a material misrepresentation to shareholders.
AI washing examples that expose the pattern
The pattern is consistent. Companies make AI claims because the market rewards the narrative. Block’s stock didn’t jump 24% because it laid off 4,000 people. It jumped because investors believed AI made the company fundamentally more efficient. Whether that’s true is a separate question.
Are tech layoffs in 2026 actually because of AI?
The numbers tell a complicated story. Challenger, Gray & Christmas tracked 12,304 U.S. job cuts citing AI in January and February 2026 alone. That’s 8% of all layoffs during that period. Since 2023, when the firm started tracking AI as a layoff reason, the total has reached 91,753 announced cuts.
But that data only tells you what companies said. It doesn’t tell you what’s true.
A paper from the National Bureau of Economic Research surveyed executives and found that 90% reported AI has had zero impact on employment at their companies over the past three years. Zero. Not “minimal.” Not “small.” None.
Goldman Sachs economists project that AI-driven displacement could add up to 0.3 percentage points to the unemployment rate in 2026. That’s meaningful in aggregate but far smaller than the Silicon Valley narrative suggests. The bank’s base case estimates the broader adoption timeline at roughly a decade, with 6-7% of workers displaced during the entire transition period.
So what’s actually happening? The layoffs are real, but the AI explanation is often a convenient story layered on top of simpler corporate math. Many tech companies doubled or tripled headcount between 2020 and 2022 chasing pandemic-era demand that didn’t last. By 2025, revenue growth had slowed, margins were under pressure, and activist investors wanted cuts. AI gave executives a forward-looking, market-friendly justification for corrections they would have made anyway.
AI washing vs greenwashing
AI washing follows the greenwashing playbook almost step for step. Both exploit public enthusiasm for something perceived as positive. Both involve companies making claims they can’t substantiate. And both are getting regulatory attention for the same reason: investors make financial decisions based on these claims.
The SEC has drawn the parallel explicitly. Its Cybersecurity and Emerging Technologies Unit treats AI washing the same way its environmental disclosure teams treat greenwashing, as a form of securities fraud when material misrepresentations are made to investors.
There’s one key difference. Greenwashing claims are often vague enough to be technically defensible (“committed to sustainability” is hard to disprove). AI washing, especially the product variety, tends to be more binary. Either your product uses machine learning or it doesn’t. Either your checkout system is automated or there are 1,000 people in India watching security cameras. That binary quality makes AI washing easier to prosecute but also easier for companies to avoid. Just tell the truth about what your product does.
How to spot AI washing before you get fooled
Whether you’re evaluating a tool for your business, assessing a company’s stock, or trying to understand why your industry is restructuring, there are reliable ways to separate real AI from marketing fiction.
Ask what the AI actually does. Legitimate AI companies can explain, in plain language, what their model is trained on, what decisions it makes, and what still requires human judgment. If the answer is vague or filled with buzzwords (“our proprietary AI engine”), dig deeper or walk away.
Check whether the “AI” existed before the hype. Many products now marketed as AI-powered were doing the same thing under a different label two years ago. If a CRM calls its search bar “AI-powered” when it’s running the same keyword matching it always has, that’s AI washing.
Follow the money on layoffs. When a company announces AI-driven workforce reductions, look at its hiring history. If headcount doubled during 2020-2022 and the cuts bring it back to pre-pandemic levels, the explanation is probably simpler than AI transformation. Also check whether the company is simultaneously spending heavily on AI infrastructure. That suggests cuts are funding AI, not being caused by it.
Look for specifics. Real AI adoption comes with measurable outcomes. Klarna reported its AI assistant handled 2.3 million customer service conversations in its first month, equivalent to the work of 700 full-time agents. That’s specific and verifiable. A company saying “AI is making us more efficient” without numbers is telling you nothing.
Read the regulatory filings, not the press release. SEC 10-K filings and earnings call transcripts contain more honest assessments of AI’s actual impact than press releases. Companies are legally required to disclose material risks and changes in their filings. The marketing department has no such obligation.
Frequently asked questions
▾ What is AI washing?
AI washing is when companies exaggerate or fabricate the role of artificial intelligence in their products, services, or workforce decisions. The term covers both marketing deception (labeling products “AI-powered” when they’re not) and employment deception (blaming AI for layoffs driven by cost-cutting or overhiring corrections). The SEC issued its first AI washing penalties in March 2024.
▾ Is AI washing illegal?
Product AI washing can violate securities law when companies make false claims to investors about their AI capabilities. The SEC has fined companies including Delphia ($225,000) and Global Predictions ($175,000), and the DOJ charged Nate Inc.’s founder with fraud for fabricating a $42 million AI shopping app. Employment AI washing is not yet explicitly illegal, but securities lawyers say it could constitute material misrepresentation if companies cite nonexistent AI efficiencies to justify layoffs to shareholders.
▾ What are the most well-known AI washing examples?
Amazon’s Just Walk Out checkout system is the most cited product example, as it relied on 1,000 workers in India rather than AI to verify 70% of transactions. In the employment category, Block’s February 2026 decision to cut 4,000 workers citing AI drew scrutiny after Bloomberg reported the cuts appeared to reverse pandemic-era overhiring. The SEC’s case against Nate Inc., where a founder raised $42 million on a fabricated 90% automation rate, is the most extreme enforcement case to date.
▾ What is the difference between AI washing and greenwashing?
Both involve companies making misleading claims to appear more innovative or responsible. Greenwashing exaggerates environmental credentials, while AI washing exaggerates AI capabilities. The SEC treats both as potential securities fraud when the claims are material to investor decisions. One key difference: AI washing claims tend to be more binary and easier to disprove (either the product uses machine learning or it doesn’t), which may lead to faster regulatory enforcement.
▾ Are the 2026 tech layoffs actually caused by AI?
Mostly not. A National Bureau of Economic Research paper found that 90% of executives surveyed say AI has had no impact on employment at their companies. Challenger, Gray & Christmas data shows AI was cited in 12,304 U.S. job cuts in early 2026, but many of those cuts reversed hiring surges from 2020-2022. OpenAI CEO Sam Altman has publicly stated that many companies are AI washing their layoffs. Goldman Sachs projects AI displacement will add about 0.3 percentage points to unemployment in 2026, far less than headlines suggest.
▾ How can you tell if a product is really AI-powered?
Ask three questions: What data is the model trained on? What decisions does the AI make versus what still requires human judgment? And can the company point to measurable outcomes? Legitimate AI products come with specifics. Klarna’s AI assistant handled 2.3 million conversations in its first month, replacing work equivalent to 700 agents. If a company can only offer buzzwords like “proprietary AI engine” without explaining what it actually does, that’s a red flag.
