CEOs using AI to cut headcount are playing the wrong game

Every week, a CEO drops the word "AI" into a layoff memo and the stock price pops. The financial press runs with it, and another data point gets added to the pile of evidence that artificial intelligence is coming for human jobs. I've watched this cycle play out enough times now that I want to push back — not with a theory, but with data.

At Gong, we have a view into something most commentators don't: millions of B2B sales interactions, captured and analyzed in real time, before the PR teams get involved and before the narratives get cleaned up for investors. What we're seeing in that data is materially different from what's being reported.

Since 2024, deals where AI is discussed have increased 85%. That number gets cited constantly as evidence of disruption. Here's the number that doesn't: deals where hiring is discussed have remained unchanged over the same period. If AI were replacing human workers at the scale the headlines imply, you'd expect those two lines to move in opposite directions with much more force. They aren't. The gap between them is, in my view, the most underreported economic signal in business right now.

The broader labor market tells a similar story. LinkedIn's data shows AI has already created 1.3 million new jobs globally, with AI engineer ranking as the fastest-growing job title for two consecutive years. Between 2023 and 2025, LinkedIn added 639,000 AI-related job postings in the U.S. alone. Meanwhile, high-skill sales roles — sales manager and sales engineer — are each projected to grow at 5% through 2034, faster than the economy overall. The pattern is consistent: AI is not eliminating the jobs that require judgment, relationships, and technical depth. It is making those jobs more valuable and more in demand.

But the finding that really changed how I think about this came from our State of Revenue AI 2026 research, which surveyed more than 3,000 revenue leaders. When we looked at the organizations with the most mature, most deeply embedded AI deployments — not the ones experimenting, but the ones actually running their revenue motion on AI — those companies reported the most aggressive hiring plans of any group in the study.

I read that as a structural argument about what AI actually does to an organization's ambition.

Radiology is the clearest real-world proof of this dynamic. For years, the conventional wisdom was that AI would replace radiologists — the work seemed formulaic, pattern-based, automatable. Between 2018 and early 2025, radiology caseloads skyrocketed 25%. Over the last decade, the number of active radiologists in the U.S. has grown by about 10%. What happened was not replacement — it was expansion. AI made imaging faster and cheaper, which made imaging more accessible, which created more demand for the humans who interpret it. The ceiling went up. That is exactly the pattern I expect to see play out in revenue organizations over the next five years.

The distinction that matters is tasks versus jobs. Tasks are being automated: drafting, logging, summarizing, routing, researching. For people whose work consisted primarily of those tasks, the disruption is genuine and the adjustment is hard. That deserves more serious attention than it typically gets in the AI optimism conversation, including from people like me. But a job is not a list of tasks. A job is a bundle of responsibilities, relationships, and judgment calls — and when AI absorbs the tasks, the humans don't disappear. They evolve.

This happens in two distinct ways. The first is capacity. AI raises the ceiling on how many accounts you can intelligently cover, how many markets you can pursue, and how many customer relationships you can manage with genuine quality. When the ceiling goes up, you need more people to go after the larger opportunity, not fewer. The unit of revenue org design stops being "how many reps do I need for X accounts" and starts being "how much surface area can this team cover with AI in the system."

The second is effectiveness. It's not just that people can take on more — it's that they act on better information at every decision point. A rep three weeks into a deal knows which move actually saves it. A leader entering a new market has a clearer signal on where to place the bet. The gap between knowing what to do and doing the right thing closes. These aren't separate levers. They compound.

When AI absorbs the drudgery, the nature of the job itself changes. The work stops being about task execution and starts being about judgment — reading a room, knowing when to push and when to listen, building the kind of trust that no automation can manufacture. AI makes capacity abundant. It makes genuine judgment and human connection the scarce, valuable thing.

Here's my actual prediction: the companies that move furthest and fastest on genuine AI adoption in the next 18 months will outperform their growth targets, and they will be net job creators. Not in spite of their AI investments, but because of them. The organizations treating AI as a cost-cutting mechanism are optimizing for the wrong variable. The ones treating it as a growth engine are building something their competitors will struggle to catch.

My challenge to other CEOs is straightforward: if you're serious about AI, measure your commitment by what you build and who you hire to pursue the opportunity it creates — not by how many people you let go.