A corporate boardroom with empty chairs around a long table and a single glowing laptop, muted colors, documentary photography
Career & Future

Andreessen Says AI Layoffs Are a 'Farce' — Companies Are 75% Overstaffed and Using AI as Cover

The AI layoff story? It's a farce, says Andreessen. Companies are 75% overstaffed from COVID bloat and using AI as a silver bullet excuse. But the workers still lose their jobs.

AI layoffsMarc AndreessenForresterworkforcecareer impact

Marc Andreessen has a message for everyone wringing their hands about AI taking jobs: it’s a farce.

In a widely-shared March interview on the 20VC podcast, the a16z co-founder didn’t mince words. The real story behind 2026’s wave of tech layoffs isn’t artificial intelligence — it’s corporate accountability for a pandemic hiring binge that went sideways.

“Essentially, every large company is overstaffed,” Andreessen said. “It’s at least overstaffed by 25%. I think most large companies are overstaffed by 50%. I think a lot of them are overstaffed by 75%.”

And the excuse? “Now they all have the silver bullet excuse: Ah, it’s AI.”


The Overhiring That Started It All

The numbers back up the bloat thesis. U.S. tech employment surged to 8.3 million by May 2020, with companies like Amazon doubling headcount from 2019-2021. Free money was flowing, interest rates were near zero, and remote work made it easy to hire fast and worry later.

Then rates went up. The cost of capital climbed roughly 5 percentage points. Suddenly, every headcount had a real price tag. Companies that had been treating engineers like collectible cards discovered they’d overpaid for a deck they couldn’t play.

The layoffs that followed — Oracle, Atlassian, Block, and dozens more cutting thousands — weren’t driven by AI capability. They were driven by a financial reality check that was always coming. AI just made a convenient headline.


The ‘AI Washing’ Problem

Andreessen’s argument aligns with what critics are calling “AI washing” — the practice of attributing routine cost-cutting to artificial intelligence to make it sound strategic rather than desperate.

As SF Standard reported, this pattern echoes across the industry. Salesforce’s Marc Benioff and Block’s Jack Dorsey have made similar observations. The 2026 layoff tracker shows roughly 214 companies and 90,000 jobs cut so far — with many press releases conveniently mentioning “AI transformation” or “efficiency gains.”

But here’s what Andreessen points out that most coverage misses: “Until literally December [2025], [AI] was not actually good enough to do any of the jobs that they’re actually cutting.” The timeline doesn’t add up. Layoffs started well before AI could meaningfully replace the roles being eliminated.


Forrester’s Regret Data Tells the Other Half

Andreessen’s “farce” framing gets support from an unlikely ally: data. Forrester reports that 55% of employers who cut jobs for AI now regret it, and many are already rehiring — often offshore, at lower wages.

This is the layoff boomerang. Companies that replaced humans with chatbots discovered what they lost: institutional knowledge, customer relationships, creative problem-solving. Now they’re paying the price twice — once for the severance packages, and again for the rehiring scramble.

The Forbes coverage of this story calls it the “dirty secret behind AI layoffs”: companies aren’t just cutting jobs, they’re often replacing them with cheaper workers elsewhere while claiming AI did the work.


Why Andreessen Is Half-Right (and Half-Wrong)

Andreessen’s “100% incorrect” framing about labor displacement is where his argument starts to fray.

Yes, many 2026 layoffs are pandemic bloat corrections. Yes, AI is a convenient scapegoat. But the “this is all just efficiency” argument ignores some uncomfortable realities:

  1. AI is starting to displace specific roles. Programming assistants, customer service bots, and content generation tools are genuinely reducing headcount in targeted functions. The Cognizant forecast of 500,000+ AI-related job cuts in 2026 may be hyped, but it’s not zero.

  2. “Overstaffed by 75%” is a VC talking point, not a management principle. If 75% of your workforce is truly unnecessary, you have a management failure, not a hiring problem. CEOs who let that happen should answer for it — not blame the market.

  3. The productivity argument cuts both ways. Andreessen says AI will create jobs through productivity gains. Maybe. But the data on who benefits from those gains suggests it won’t be the same people who lost their jobs.

  4. Offshoring is the real story. When Forrester’s regretful employers rehire, they’re often hiring in India, the Philippines, or Eastern Europe. The “AI displacement” narrative obscures a global labor arbitrage that’s been accelerating for decades.


The NZ Angle

New Zealand tech workers are watching this unfold in real time. Our tech sector is smaller and more exposed — when global companies cut, the local offices feel it first.

But there’s a specific local wrinkle: NZ companies didn’t overhire as aggressively during COVID. Our bloat is smaller. Which means if AI starts genuinely displacing roles — not just providing cover for cost-cutting — we’ll feel it faster because we have less fat to trim.

The real question for NZ workers isn’t whether Andreessen is right about AI being a scapegoat. It’s whether anyone is preparing for the scenario where he’s wrong.


🔍 THE BOTTOM LINE

Andreessen’s “farce” take is a useful corrective to the panic narrative, but it’s also self-serving from a venture capitalist whose portfolio benefits from the AI-hype cycle. The truth is messier: some layoffs are pandemic bloat, some are genuine AI displacement, and a disturbing number are old-fashioned offshoring wearing a silicon mask. The workers who lost their jobs don’t much care which category they fall into. They still need to pay rent.


SOURCES

Sources: Fortune, Yahoo Finance, SF Standard, 20VC Podcast, Forbes