The AI ROI Gap: Why CFOs Must Move Beyond the Experimentation Phase

**Opinion:** The era of AI 'experimentation' as a valid budget line item must end. For too long, enterprise AI initiatives have been measured by vanity metrics rather than concrete ROI. If a tool does not demonstrably increase revenue per worker or expand profit margins, it is a costly hobby, not a strategic asset.

As TechCrunch first reported, this skepticism is now being quantified by Torsten Slok, chief economist at Apollo Global Management. In a recent note, Slok stated that artificial intelligence has provided "zero financial payoff to corporate America outside of tech."

To reach this conclusion, Slok analyzed the "S&P 493"—the S&P 500 excluding the seven dominant tech giants: Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla. He found that for these non-tech corporations, profit margins and revenue per worker have remained flat, suggesting that massive capital expenditures are not yet driving productivity gains.

Slok argues that the primary hurdle is the distinction between calculation and judgment. While AI is an effective calculator, Slok posits that business success is driven by human elements that AI cannot grasp, including accountability, trust, corporate culture, unwritten rules, ethical trade-offs, and fairness. Without these, the ability of an LLM to impact the bottom line remains limited.