AI Promised Efficiency. Workers Want Higher Pay Anyway.

The pitch was elegant in its simplicity: deploy artificial intelligence, eliminate redundancy, boost productivity, justify wage freezes. By late 2024, it had become the corporate equivalent of a get-rich-quick scheme, except the only ones getting rich were the consultants selling the dream.

But something unexpected happened on the way to the lean, mean, AI-optimized future. Workers didn't disappear. Labour markets tightened. Wage pressure persisted. And the promised productivity gains—the ones that were supposed to make salary restraint not just acceptable but inevitable—remained stubbornly elusive.

The numbers tell a story of widening gaps between narrative and reality. Across major economies tracked by the OECD, wage growth acceleration has outpaced productivity gains in 19 of the last 24 months, even as AI adoption rates in enterprise environments reached 45 percent by Q3 2024. Translation: companies are investing heavily in the technology while workers are increasingly demanding compensation that reflects the tightness of the labour market, not the promise of algorithmic efficiency.

In the US, despite 260,000 tech sector layoffs announced in 2023-2024, unemployment remains below 4 percent. The Netherlands reports similar pressure, with tech roles commanding 12-15 percent wage premiums over pre-pandemic levels despite widespread AI integration projects. South Korea, where Samsung and LG have heavily promoted automation narratives, is experiencing the highest wage growth in manufacturing since 2008. Germany's labour shortage in skilled sectors has actually worsened, not improved, even as DAX companies announce major AI investments.

The disconnect is not subtle. Companies frame AI deployment as a cost-containment strategy. Workers read it as: management has decided our roles are expendable, which means we need to extract maximum compensation while we still have leverage. One cancels out the other.

Consider the stated case: McKinsey's 2024 report claiming AI could drive a 1.4 percent annual productivity increase globally. Sounds significant until you factor in actual output metrics from companies that have deployed AI at scale. A manufacturing firm in the EU reported to Deloitte that their AI-powered quality control system reduced defects by 23 percent—but required three additional human supervisors to maintain, recalibrate, and handle edge cases. A London-based financial services firm saw their robo-advisory platform handle 40 percent more client interactions but couldn't reduce headcount because compliance and relationship management remained stubbornly human-dependent.

This is the productivity paradox that CFOs dare not mention in earnings calls: AI doesn't eliminate work so much as it redistributes it, often to roles that require higher wages and more sophisticated skill sets. The narrative of "do more with less" crashes into the reality that you still need someone to teach the AI, someone to catch its mistakes, and someone to explain to a customer why an algorithm made a decision that destroys shareholder value.

Meanwhile, workers have internalized the lesson: if management is serious about needing fewer of us, individual compensation needs to rise. Across Europe and North America, retention demands—not requests, demands—have jumped 34 percent since Q1 2024 according to recruitment data from Robert Half. In Singapore and Hong Kong, where tech talent is most scarce, companies are offering signing bonuses that would have seemed absurd three years ago, even as they simultaneously announce "AI-driven efficiency initiatives."

The bluff is being called because labour markets have memory. Workers know that layoffs are often about strategic repositioning, not genuine productivity gains. They know that companies announcing AI investments aren't necessarily expecting to do more with less—they're preparing for a future where the people who remain will be more valuable, not less. So they're positioning themselves to capture that value now, before the winnowing happens.

This creates a peculiar moment for corporate leadership. The AI narrative was supposed to buy time, justify restraint, and create a rational case for lean operations. Instead, it's accelerated the very wage pressure it was designed to contain. The technology is real. The productivity gains are real but constrained. But the ability to use "we're investing in AI" as a reason not to raise salaries? That's evaporating fast.

Q3 earnings will be interesting. Not because AI adoption will suddenly produce the promised 30 percent efficiency gains—it won't. But because the conversation is shifting from "AI will save us money" to the messier question of what happens when the efficiency narrative collides with workers who've decided they're not waiting around to find out if the cuts are coming.