AI delivers ‘zero financial payoff’ to most companies, Apollo economist says

October 25, 2026 · DigitalEdu

What Happened

A recent analysis by Apollo Global Management’s chief economist, Torsten Sløk, has ignited debate among market analysts and corporate strategists. Sløk’s assessment cuts through the prevailing optimism: despite billions of dollars funneled into artificial intelligence infrastructure and the integration of machine learning tools across diverse sectors, the technology has yet to translate into measurable financial returns for the majority of corporate America. The statement, delivered during a market outlook presentation, frames the current AI spending spree as an investment without a confirmed dividend. For a landscape where AI adoption is nearly ubiquitous, the implication is stark—the much-touted productivity revolution is, as of yet, failing to show up on the balance sheet.

Why It Matters for Businesses

The disconnect between massive capital expenditure and tangible profit is a critical concern for business leaders navigating the current economic climate. When a technology is billed as a general-purpose transformative force, the absence of widespread financial payoff suggests that adoption is not synonymous with optimization. For non-tech firms, the challenge lies in moving beyond pilot programs and experimental implementations to integrate AI in ways that directly impact revenue generation or cost reduction. The "zero payoff" observation serves as a cautionary signal that simply deploying AI tools is not a strategy; without a clear linkage to business outcomes, the technology risks becoming an expensive overhead rather than an asset.

What To Watch

Analysts and investors will likely be monitoring two primary fronts in the coming quarters. First, the pace at which companies move from experimental AI usage to mission-critical applications that drive specific key performance indicators. Second, the evolution of AI pricing models and the development of industry-specific tools that lower the barrier to entry for non-technical sectors. There is also attention on whether the productivity gains, which may be materializing in specific pockets of the economy, are being offset by the high costs of talent acquisition and infrastructure maintenance. The trajectory of these factors will determine if the current investment cycle yields returns or corrects toward more sustainable spending levels.

The Bottom Line

For business owners, the takeaway is a need for rigorous ROI assessment. The Apollo economist’s warning underscores a fundamental truth: AI is a tool, not a guaranteed profit center. Companies must move beyond the hype and focus on concrete use cases that align with their specific operational goals. Until the link between AI deployment and financial performance is clearly established, the prudent approach is measured investment, targeted implementation, and a relentless focus on how these technologies actually solve business problems rather than merely following the industry trend.

Source: Original Article

Related: Block’s new voice model sounds eerily human, Forced to pick, Grok’s AI model chooses Republicans. OpenAI’s picks Democrats. , OpenAI says it will no longer train its AI models on the live internet, citing “significant risks”.

For more on this, see Alibaba tests new business model for Qwen open-source AI.

What happened: TechCrunch reports that Apollo Global Management's chief economist Torsten Sløk says AI has delivered zero financial payoff to most companies despite massive investment. The analysis argues adoption is not the same as optimization.

What this means if you run a small business

If you own a service firm with 10 to 30 employees, you have likely bought AI writing, coding, or meeting tools. The warning means those subscriptions may not show up as higher margin or lower headcount unless you change how work actually flows.

If you run a light-manufacturing or logistics outfit with 20 to 50 staff, you may have piloted predictive maintenance or routing software. The finding suggests pilots stay pilots until someone owns the handoff from model output to daily decision.

Retail and hospitality operators with 5 to 20 people often add chatbots or scheduling assistants. The report implies those tools cut cost only when they replace a task completely, not when they just add a layer of review.

What it would take to act on this

Map one repeatable process end to end, identify the step where a human currently decides, and test whether an AI output can replace that decision without adding review time. Assign one person to own the metric that proves the swap worked. The cost depends on scope.

Our take

The headline is a useful corrective: buying tools is not a strategy. Small businesses that treat AI as a process-redesign lever, not a software purchase, are the ones most likely to see a return.

Questions this raises

How do I know which of my current AI subscriptions are actually wasting money?
Run a 30-day log of time spent prompting, correcting, or supervising each tool; if the net time saved is zero or negative, cancel it.

What is the smallest process I can redesign this quarter to test whether AI can replace a decision?
Pick a single, high-volume, rules-based decision — like quote approval or shift assignment — and build a measurable before-and-after comparison.