You’ve seen the numbers. Maybe you’ve felt a chill reading them.
“Company X cuts 10,000 jobs, cites AI efficiency.” Another headline. Another wave of anxiety ripples through your organisation. Another executive somewhere asks, “Should we be doing this too?”
I’ve been tracking these announcements for two years now. Every press release, every interview, every news story where a company publicly declares they’re replacing humans with artificial intelligence. The running total? 217,054 jobs globally.
That number should terrify you. But here’s the thing: the number is lying to you. Not completely. But enough that you need to understand what’s really happening before you make decisions that affect real people’s lives.
The Companies Selling AI Are the Ones “Proving” It Works
73.3% of all AI displacement announcements come from companies that sell AI products or services.
That’s 159,089 jobs announced as AI replacements—by the very companies who directly profit from convincing you that AI can replace your workforce too.
Amazon. Microsoft. IBM. Intel. These aren’t just laying people off; they’re selling you the tools to do the same. Then you’ve got Accenture and TCS, whose consulting arms are absolutely swimming in “AI transformation” revenue. SAP and Salesforce wedged AI into their enterprise platforms faster than you can say “quarterly earnings call.”
Think about that for a second. If you sold gherkins, wouldn’t it be convenient to announce that you’d replaced your entire maintenance team with gherkins? “Look how effective gherkins are! We use them ourselves! Now, would you like to buy some gherkins?”
That’s funny, isn’t it? The primary evidence that AI is replacing jobs comes from companies whose business model depends on you believing AI replaces jobs.
I’m not saying these companies are lying. I’m saying their incentives make the data impossibly murky. These are mature organisations that would be going through restructuring regardless. They’d be trimming headcount as they shift from growth mode to margin optimisation. That’s just how business cycles work.
But slapping “AI” on the press release? That’s not a workforce strategy. That’s marketing.
The Valuation Game Makes It Even Murkier
Some companies announce AI layoffs because it makes them look smart to investors, not because the technology actually did anything.
There’s a smaller cluster in my data that caught my attention: Klarna, xAI, Scale AI, Fiverr, Chegg. At the time of their announcements, every single one had a pressing need to look lean, efficient, and technologically sophisticated. They were either approaching IPOs, seeking funding rounds, or desperately trying to justify their valuations.
When you’re trying to convince venture capitalists that you’re worth billions, “we’ve achieved operational efficiency through cutting-edge AI” sounds a lot better than “we overhired during the boom times and now we’re paying for it.”
Here’s the uncomfortable truth I’ve learned from years of engineering leadership: press releases are fiction written for shareholders. The internal reality is almost always messier. The “AI efficiency” might be a spreadsheet macro. The “automation” might be three interns and a Python script. The “transformation” might be management finally noticing that half the team was doing work that shouldn’t have existed in the first place.
I’ve been in rooms where leadership discussed how to frame necessary layoffs. The pressure to attach a trendy narrative is immense. “We’re not failing to manage growth—we’re embracing AI!” It’s not malice. It’s survival instinct in a market that punishes honesty and rewards buzzwords.
So when you see these headlines, ask yourself: who benefits from me believing this story?
Real Displacement Is Happening—Just Not Where You’d Expect
About 20% of the announced layoffs—roughly 43,000 jobs—appear to be genuine AI displacement, mostly in organisations with no incentive to spin the narrative.
This is the part that actually matters. Strip away the marketing and the valuation games, and you find the banks: Wells Fargo, ABN Amro, DBS, Westpac. The traditional industries: Nestlé, Lufthansa. The governments: Uzbekistan’s civil service, the US GSA, the Pentagon’s Defence Technical Information Centre.
These organisations don’t sell AI. They’re not chasing venture capital. They have no reason to lie about why they’re cutting jobs.
And when you look at what they’re automating, it’s almost boringly predictable: document processing, basic customer service triage, data entry, routine compliance checks. The kind of work that was always going to be automated eventually. AI just accelerated the timeline.
This is real. It’s happening. And if you’re a leader, you need to be honest with your teams about it.
But here’s what the doom-mongers miss: 43,000 jobs over two years, across all these organisations globally, is not the robot apocalypse. It’s technological change doing what technological change has always done—eliminating routine tasks while creating demand for people who can do what machines can’t.
The question isn’t whether AI will change work. It will. The question is whether you’ll be the leader who uses that change as cover for poor decisions, or the one who navigates it honestly.
The Bubble Will Burst. The Technology Won’t Disappear.
We’re in a bubble. The comparisons to Railway Mania and the dot-com crash aren’t hyperbole—the patterns are nearly identical. A transformative technology emerges. Capital floods in. Valuations detach from reality. Eventually, something breaks.
But here’s what the bubble-callers forget: we still have railways. We still have e-commerce. The investors who bought at the peak got destroyed. The grifters and hype merchants got exposed. But the people who quietly built useful things with the underlying technology? They emerged from the rubble and kept building.
That’s who benefits from these shifts. Not the companies screaming about AI transformation while selling you transformation services. Not the startups inflating layoff numbers to impress VCs. The winners are the boring ones: the teams that looked at a real problem, asked whether AI could help solve it, and then actually tested the hypothesis before betting people’s livelihoods on the answer.
In a world that seems allergic to sensible middle ground, your job is to find it anyway. Neither AI hype merchant nor AI denier. Just a leader who asks hard questions and refuses to sacrifice people on the altar of a press release.