Torsten Sløk: ADP Data Shows AI Builds Jobs, Not Cuts
Key insights
- Apollo chief economist Torsten Sløk cites ADP employment data as evidence AI is creating, not eliminating, jobs.
- In 2025, AI-attributed U.S. layoffs totaled roughly 55,000, just 4.5% of 1.2 million total layoffs that year.
- Block and Klarna both cite AI when cutting staff, but Yale Budget Lab and Forrester call displacement fears largely speculative.
Why this matters
Sløk's ADP-backed argument hands CFOs and boards a data point to resist AI-driven headcount reduction pressure at a moment when Block, Klarna, and Amazon are publicly attributing workforce cuts to AI efficiency. The 4.5% figure places AI-attributed layoffs as a small fraction of total U.S. job loss in 2025, directly challenging narratives that AI is structurally reshaping employment at scale. For technical leaders, the salary pressure Sløk identifies in data center hiring signals where AI talent competition is concentrating now: infrastructure roles, not application-layer positions.
Summary
Torsten Sløk, Apollo's chief economist, uses ADP data and the Jevons Paradox to argue there is 'zero evidence' AI destroys jobs: cheaper technology creates demand, not unemployment.
Firms are actively hiring AI implementation experts while data center expansion raises salaries for AI specialists and prices for semiconductors and energy.
Essentially: (Sløk, Apollo) vs. (Anthropic CEO, Amazon's Jassy) on whether AI destroys or grows headcount.
- Block cut roughly 4,000 employees, nearly half its workforce.
- Klarna shrank from 4,500 to 3,500, saving $10 million annually.
- 55,000 AI-attributed layoffs in 2025 were just 4.5% of 1.2 million U.S. total.
Yale Budget Lab and Forrester say AI job loss fears are 'largely speculative,' though Klarna's CEO warns a 'massive shift' is coming.
Potential risks and opportunities
Risks
- If Block's roughly 4,000-person reduction and Klarna's 4,500-to-3,500 cut are followed by similar AI-attributed layoffs across fintech and enterprise software, Sløk's 'zero evidence' framing accumulates contradictions by year-end 2026.
- Sam Altman acknowledged companies may be 'blaming AI for layoffs that they would otherwise do,' meaning ADP data Sløk relies on could be distorted by misattributed layoffs, undermining the empirical foundation of his argument.
- Anthropic CEO's claim that AI 'could eliminate 50% of white-collar jobs,' if it materializes even partially, would directly invalidate Sløk's current position and expose Apollo Global Management to reputational risk in the AI employment debate.
Opportunities
- AI implementation staffing and consultancies benefit directly from the Sløk-documented hiring wave for AI implementation experts, particularly at firms undergoing data center expansion.
- Firms already scaling data center infrastructure can use Sløk's data to justify accelerated AI talent acquisition before salary pressure for AI specialists and semiconductor professionals becomes more acute.
- Forrester, Yale Budget Lab, and the National Bureau of Economic Research are positioned to set the authoritative measurement framework for AI-attributed layoffs, giving them significant influence over the policy and investment debate Sløk is engaging.
What we don't know yet
- Whether ADP employment data controls for the lag between AI implementation and downstream job elimination in affected white-collar sectors
- How long Sløk's 'AI implementation experts' hiring wave persists before those roles themselves become targets for automation
- Whether Klarna's $10 million in annual AI savings has accelerated further headcount reductions beyond the 4,500-to-3,500 cut that completed in 2024
Originally reported by fortune.com
Read the original article →Original headline: Apollo Chief Economist: 'Zero Evidence' AI Is Killing Jobs — Cites Jevons Paradox and ADP Hiring Data