Apollo's Chief Economist Says There Is 'Zero Evidence' of AI-Related Job Losses
As headlines continue to scream about artificial intelligence replacing human workers, one of Wall Street's most closely followed economists is pushing back hard. Torsten Sløk, chief economist at Apollo Global Management, published a blog post stating there is "zero evidence of job losses because of AI." His remarks, backed by real employment data, sparked a wave of reaction across the tech and finance world — and raised a critical question: Are fears about AI-driven unemployment overblown?
What Torsten Sløk Actually Said
In his post, Sløk cited the ADP National Employment Report — one of the most reliable private-sector employment trackers in the United States — to support his claim. Rather than seeing mass displacement, the data points to something far more nuanced: companies are not cutting workers because of AI. They are actively hiring workers who know how to use it.
"Many firms are hiring AI implementation experts, and the data center buildout is putting upward pressure on salaries for AI experts and on prices of semiconductors, equipment, and energy," Sløk wrote. "The bottom line is that the AI spending boom is stoking both employment and inflation."
This was not a one-off observation. Back in April, Sløk made a similar argument, writing that "cheaper inputs don't shrink industries. Instead, AI is going to increase both productivity and employment." That framing echoes what economists call the Jevons Paradox — the idea that as technology makes something more efficient or affordable, consumption of that resource actually increases rather than decreases.
The ADP Data: What It Tells Us
The most recent ADP National Employment Report referenced by Sløk found that private companies added nearly 110,000 jobs in the most recent reporting period. That figure, while lower than some analysts had hoped, does not reflect any systemic contraction driven by AI automation. In fact, the sectors showing the most growth include areas directly tied to AI infrastructure — technology services, cloud computing support, and data center operations.
These numbers matter because they come from actual payroll data, not surveys or projections. When economists want to understand what is truly happening in the labor market, hard payroll data is the gold standard. And by that measure, the AI apocalypse simply has not arrived.
But What About the CEOs Citing AI in Layoffs?
Here is where the picture gets complicated. While Sløk's data shows no broad labor market contraction attributable to AI, a growing number of high-profile companies have explicitly cited artificial intelligence as a reason for workforce reductions in 2025. At least a dozen major corporations have pointed to AI-driven efficiency gains as justification for laying off workers this year alone.
Notable examples include technology firms and financial services companies that have reduced headcount in specific departments — particularly in areas like customer support, content moderation, data entry, and certain tiers of software development. These are real job losses affecting real people, and dismissing them entirely would be misleading.
So how do we reconcile Sløk's "zero evidence" claim with the very real layoffs being attributed to AI? The answer likely lies in scale and replacement rate. When one company cuts 200 customer service roles and attributes the decision to AI chatbots, that loss is real — but if the broader economy simultaneously creates 1,000 new roles in AI implementation, infrastructure, and adjacent fields, the net employment effect may still be positive. Macroeconomic data captures the aggregate; individual workers feel the disruption.
AI Is Creating Jobs, Not Just Eliminating Them
One of the most important trends emerging from the current AI cycle is the explosion of entirely new job categories. Consider the following roles that barely existed five years ago and are now in high demand:
- AI implementation specialists who help businesses integrate large language models into their workflows.
- Prompt engineers who craft, test, and optimize inputs for generative AI systems to maximize output quality.
- AI ethics and compliance officers responsible for ensuring that automated systems meet regulatory and ethical standards.
- Data center technicians and engineers supporting the massive physical infrastructure required to run AI models at scale.
- AI trainers and annotators who generate and label the high-quality datasets that machine learning models depend on.
Beyond these specialist roles, companies across virtually every sector are now prioritizing AI literacy when hiring for traditional positions. Marketing managers who understand generative AI tools, accountants who can work with AI-driven forecasting models, and HR professionals who can evaluate AI-assisted recruitment software are all seeing improved hiring prospects compared to peers without those skills.
The Inflation Angle: An Underreported Consequence
Sløk's commentary also introduced a dimension that rarely makes it into mainstream AI coverage: inflationary pressure. The AI spending boom is driving up demand for semiconductors, specialized hardware, energy, and highly skilled human labor. This demand-pull dynamic is contributing to wage inflation in the technology sector, which in turn ripples outward into broader price indexes.
Data centers require enormous amounts of electricity. GPU shortages continue to affect supply chains. Compensation packages for machine learning engineers and AI researchers have reached levels that would have seemed extraordinary just a few years ago. All of this suggests that AI is not deflationary in the near term — it is, paradoxically, adding upward pressure to costs even as it promises future efficiency gains.
What Workers Should Actually Do Right Now
If the data is correct and AI is expanding the labor market rather than contracting it, the practical implication for workers is clear: the path forward runs through AI fluency, not away from it. Professionals who invest in learning how AI tools function, how to integrate them into existing workflows, and how to critically evaluate their outputs will be far better positioned than those who avoid the technology out of fear.
This does not mean every worker needs to become a machine learning researcher. It means developing enough working knowledge of the AI tools relevant to your field to remain competitive, productive, and valuable in a market that is rapidly repricing those skills upward.
The Bottom Line
Torsten Sløk's "zero evidence" statement is bold, but it is grounded in real data. At the macroeconomic level, AI has not triggered the mass unemployment wave that many predicted. It is instead fueling a hiring surge in new categories, pushing salaries higher for skilled workers, and adding complexity to an already tight labor market.
That does not mean individual disruption is not real, or that certain roles won't continue to be automated. But the aggregate picture, at least for now, looks far less catastrophic than the narrative suggests. The smartest response to AI is not panic — it is preparation.