AI Fear Has Become a Sales Pitch

AI Fear Has Become a Sales Pitch

Eleven hundred people laid off, and a record quarter announced the same week. That was Cloudflare in early May 2026: roughly 20% of its workforce cut, while the company posted $639.8 million in quarterly revenue, up 34% year over year (TechCrunch, May 8, 2026). The internal memo says nothing about financial trouble. It talks about AI: internal agent usage reportedly jumped 600% in three months.

As economic logic, it does not hold up. A company breaking its own revenue records does not cut a fifth of its staff to survive. If the financial argument fails, we are looking at something else. Not an accounting decision. A narrative.

The Story That Pays

Since 2021, the people running AI companies have described a future where human labor becomes incidental. OpenAI chief Sam Altman wrote in March 2021 that the price of labor would "fall toward zero." Elon Musk, at the London AI safety summit in November 2023, predicted "a point where no job is needed." Dario Amodei, who leads Anthropic, spoke in May 2025 of a "white-collar bloodbath," with half of entry-level office jobs at risk and unemployment possibly hitting 20%.

These three voices share a point rarely flagged: they run or own the companies that sell AI. And the value of those companies depends directly on belief in their prediction. Anthropic was worth roughly $61 billion in early 2025. A year later, its valuation on secondary markets tops $1 trillion (Entrepreneur, May 2026). Multiplied more than fifteenfold, on the promise of a tool that can replace human work.

This is where HR professor Jean Pralong, in an op-ed published by Le Point on May 13, 2026, offers a useful reading: what if the announcement of a job apocalypse is itself a product? His thesis comes with a caveat. Pralong is a human-resources specialist, not a market analyst. The link between "valuation indexed to fear" is an interpretive frame, not a demonstrated financial mechanism. But as a frame, it lights up the loop well: the more frightening the prophecy, the more capital it attracts, the more that capital funds the tools, the more the tools make the prophecy credible.

"Attributed to AI" Is Not "Caused by AI"

The numbers seem to prove the prophets right. In the first quarter of 2026, the tech sector eliminated 78,557 positions, of which nearly 48% were explicitly tied to AI and automation, according to data from the firm Challenger reported by Tom's Hardware. In April, AI became the leading stated reason for layoffs for the second month in a row (CBS News, April 2026).

One word changes everything in those statistics: "attributed." These numbers measure what companies declare, not what actually happened. And Altman himself has acknowledged the existence of "AI washing": firms file under the AI label cuts they would have made anyway, because of post-Covid overhiring or an ordinary slowdown. Announcing a layoff "because of AI" rather than "because of a management error" flatters leadership and sends markets a signal of modernity.

Meta offers the clearest illustration. The group announced 8,000 job cuts, effective from May 20, 2026, presented as a reallocation toward artificial intelligence. At the same time, its 2026 AI budget is climbing toward $135 billion, and its 2025 annual revenue topped $200 billion for the first time (The Next Web, May 2026). Nobody here is laying off to avoid bankruptcy. The cuts come amid record profitability, and they get called AI.

What Fear Makes Possible

The apocalyptic narrative has an effect long before the technology proves anything. For management, "AI will replace everything" is a convenient tool. It justifies cuts without having to admit a strategic mistake. It disciplines the teams that remain, now expected to prove their "added value against AI." And it weighs on wages: it is hard to negotiate a raise when you are repeatedly told your job is living on borrowed time.

It works a bit like a landlord telling an entire building the neighborhood is going to be demolished. Even if the demolition never happens, nobody asks for repairs anymore, and rents get renegotiated downward. The rumor does the work on its own.

What This Does Not Say

There is a trap to avoid here, and it mirrors the first one. Saying that AI fear is a sales pitch is not the same as saying AI is harmless for employment. Job displacement is real. Some tasks are already automated, some occupations are being reconfigured, and that movement will continue. The problem is not the existence of the phenomenon, it is its speed and scale, systematically inflated by those who benefit from people believing in it.

The real transformation is slow, partial, untidy. It blends into the business cycle, interest rates, post-Covid corrections. Separating it from the noise takes work that sensational headlines have no reason to do. On declic.media, we have already looked at how AI slices jobs into tasks rather than eliminating whole positions, and at how courts are starting to scrutinize layoffs justified "by AI." The picture there is considerably less clean than the slogan.

One recent detail says it all. In May 2026, Dario Amodei, the man behind the "white-collar bloodbath," began walking back his own message by invoking the Jevons paradox: AI would ultimately create more jobs than it destroys (Fortune, May 5, 2026). The prophecy adjusts when it suits the prophet.

An analyst describes what they observe. A salesperson describes what serves their product. When the two are the same person, the diagnosis stops being a diagnosis.