The Self-Fulfilling AI Layoff Prophecy
Work Forward is a newsletter dedicated to making work better for people.
Like what you’re reading? Please like ❤️, subscribe 📨, and share 🔄
Is AI killing jobs? Ask a CEO and you’ll get answers ranging from “net positive” to “30% unemployment in five years.” Ask the data and you’ll get a different story entirely.
Over the past few months, I’ve had a steady stream of conversations with executives about AI at work. The landscape changes so fast that anyone claiming to have all the answers is blowing smoke. But too often, busy executives only read headlines.
Fear as a motivator
Did AI directly enable job losses in 2025? Based on data from Brookings Institute and Budget Lab at Yale, Nick Bloom at Stanford and others, the broad answer is no. A limited number of jobs were lost to AI, and about the same number created.
As Molly Kinder at Brookings put it, “the labor market as a house is not on fire” — though there may be “lightning strikes” in early career jobs in software development, customer service, and creative work.
But headlines aren’t just clickbait, it’s still “AI” at the heart of a lot of job market weakness.
One factor? Employee motivation. As Fortune CEO Diane Brady notes, CEOs tie layoffs to AI to motivate remaining employees to adopt the technology. It’s what passes for inspiration these days: fear instead of hope and opportunity.
But it’s not just about employee motivation. AI gets mentioned to signal efficiency to Wall Street, even when companies are just cutting heads.
Fear drives the big boss as well: in a BCG survey of 604 CEOs, 50% believe keeping their job depends on getting AI right by 2026. Their pressure is becoming their employees pain.
Yeah, but what about college grads?
The impacts are accelerating. Last summer, Stanford’s Erik Brynjolfsson found a 13% decline in jobs for college graduates in customer service or computer science driven by AI. Three months later, he reported that the number was now 16%.
Unless Brynjolfsson is wrong. Google economists just torched his methodology in a new paper by Google economists Zanna Iscenko and Fabien Curto Millet. One example: Half the “AI job losses” appeared by June 2023, only six months after ChatGPT launched.
“[W]ithin a mere six months of a consumer-facing chatbot’s launch, firms across the economy not only decided that AI could replace junior staff but also managed to implement the necessary technological infrastructure, redesigned complex workflows, ensured robust data security, and executed these staffing changes at a national scale. Such a rapid and widespread operational transformation seems implausible.”
So why the Tech layoffs? While some might be a post-Covid hangover, the biggest driver is the need to hit quarterly earnings goals, covering increasing investments in AI by cutting other costs (read: humans). Another factor among the Big Tech arms merchants ismarketing: If you’re spending tens of billions on AI, you need proof it saves money somewhere. Your workforce is the demo.
As economist Robert Armstrong put it in the FT:
“Putting layoffs down to AI sounds better than saying ‘we need to keep margins high so we’re sacking some low performers’ and politically safer than saying ‘unpredictable Trump tariff policy means we are hiring less young people.’”
🔆 Special Offer 🔆
Join me at Charter’s Leading with AI Summit on Feb 10th in NYC or Feb 24 in SF to hear from a massive range of researchers and leaders who are wrestling through the “messy middle” of AI transformation.
Subscriber special: 50% off in-person registration (pending approval) with the code FORUM!
Virtual attendance is free!
Self-fulfilling prophecy
Eventually, this becomes self-fulfilling. Two-thirds of CEOs told the WSJ in December that they’re freezing hiring because of “looming uncertainty” around the economy and AI.
The belief in AI efficiency may be outstripping reality, but that’s not preventing layoffs. Professor Tom Davenport and Laks Srinivasan surveyed 1,006 global executives:
“A majority have already made either low to moderate (39%) or large (21%) headcount reductions in anticipation of AI. Another 29% is hiring fewer people in anticipation of future AI.
Only 2% have made large reductions related to actual AI implementation.”
But the job losses and weakening markets are real. We’ve moved from “low hire, low fire” to “no hire, low fire” and now predictions that range from dour to dire.
At Davos, predictions ranged from “probably about 10% displacement, but net addition” (IBM’s Arvind Krishna) to “possible 20% to 30% unemployment over the next two to five years” (Verizon’s Dan Schulman).
Combine those headlines with inflation, civil unrest, and threats to 80-year-old economic alliances, and you get the predictable nose-dive in consumer sentiment. As The Conference Board reported in late January: “Confidence collapsed to its lowest point since 2014, surpassing pandemic depths.”
(Footnote: this is the economy the WSJ called “remarkable” heading into 2025.)
Declining sentiment dampens consumer spending, long the big pole in the US economic tent. Political unrest might further cut spending through movements like Resist and Unsubscribe, aimed at curtailing tech spending to pressure business leaders.
General economic sentiment declines, people pull back, and jobs decline: it starts to sound like a recession. In past recessions, automated manufacturing jobs never returned. If the same holds for office workers, we’ll see even broader hollowing out of the middle class.
Data: Bureau of Labor Statistics; Chart: Axios Visuals
Focus on growth
There’s opportunity to change direction. Much remains in executives’ hands. Fortunately, fiduciary responsibility is a lever we can pull.
Using AI just to cut costs is a long-term dead end. You get more engagement pointing to outcomes that grow the business and skills than saying “efficiency” again.
Data backs this up. In a PWC survey, more executives used AI to drive revenue growth (29%) than cost cuts (26%). The holy grail of efficient growth is revenue up and costs flat or down: 20% are hitting it.
Source: PWC January 2026
The key, as noted by a group of execs in a recent Charter dinner, is understanding the long game.
“We are not going to get disintermediated by somebody who’s 10% more efficient,” one executive told me. “We are going to get disintermediated by a 10-person team who fundamentally thinks about how a company can be built differently.”
That 10-person team isn’t optimizing the old model to get 10% better. They’re building something entirely new while you’re still debating headcount. Think they’re not a threat? Ask the enterprise software industry this week: My former employer, Salesforce, is down 43% over the past year.
Intervention needed
As interviewer Nathan Gardels points out in his conversation with MIT economist David Autor: “Productivity growth and wealth creation are being divorced from jobs and income — that is the central social challenge. Increasingly, the gains will flow to capital and decreasingly to labor.”
The overlords of our robot overlords will get very rich while the middle class gets hollowed out. How far can an already-expanding K-shaped economy stretch? Not something I’d care to find out.
Brookings’ Molly Kinder points out solutions start with how we measure AI:
“Every time we’re measuring AI, it’s whether it’s better than a human. Why are we trying to best humans? Why isn’t the benchmark making the human better?”
She suggests redirecting federal innovation funding: “Why are we not steering that toward a new benchmark where you prove the output levels up humans?”
Even Jamie Dimon knows massive unemployment is bad for society and business. At Davos, he discussed guardrails: employment incentives, possibly bans on mass AI replacement.
Policy makers can reshape incentives through tax breaks for employing younger people, education investment accounting, and funding allocation. But I have grave doubts when it comes to policy makers in the US these days.
Who will break from the pack?
Following the rest of the herd has never generated massive, outsized wins, but that’s where we are today. Perhaps some of the 50% who aren’t fearing for their jobs in 2026 will take a different direction.
The leader who gets off the efficiency bandwagon and focuses on transformation — one that’s as much about their people as their profits — has a massive advantage. People will flock to them. The leader who says “we’re going to drive radical growth, and that means all of you have massive opportunities ahead if you lean in with me” will win.
Unheard of? Ask just-now-former Walmart CEO Doug McMillon, who said they would create as many jobs as they eliminate:
“What we want to do is equip everybody to be able to make the most of the new tools that are available, learn, adapt, add value, drive growth—and still be a really large employer years from now.”
Whether it’s researchers, policy makers, or corporate leaders, we need to advocate for better solutions. But business leaders control the levers that matter most right now.
The time for these conversations was last year. Now we’re racing the clock.
Hot takes, think tank pieces and wild ideas welcome: what’s yours?
Good insight, Brian. Can we translate part of your article into Spanish with links to you and a description of your newsletter?