The AI revolution promises unprecedented efficiency and growth, but who is actually profiting? A major new assessment from Apollo Global Management reveals a stark divide, with tech giants reaping rewards while most other firms see little financial payoff.
The Great AI Divide
A recent analysis by Torsten Slok, chief economist at investment firm Apollo Global Management, has sent a sobering message through the business world: despite massive investment and hype, AI has so far delivered 'zero financial payoff' to companies outside of the tech sector. The report examines the S&P 500 by splitting it into two groups: the 'Magnificent Seven' tech giants (like Alphabet, Microsoft, and Nvidia) and the remaining 493 companies. The findings are clear: key metrics like revenue per worker and profit margins have soared for the tech titans since the AI boom began. For the rest of corporate America, those same figures have remained almost entirely flat, suggesting the promised productivity miracle has yet to materialise on their balance sheets.
Why Tech Giants Are Winning
The advantage for tech giants is structural. Many of them are not just using AI, they are building and selling the foundational infrastructure that powers it. Companies like Microsoft, Google, and Amazon benefit from owning the cloud platforms where AI models are trained and run. Chipmaker Nvidia has seen explosive growth by providing the specialised GPUs essential for AI development. This creates a virtuous cycle: they profit from the widespread corporate investment in AI, while also leveraging the technology to enhance their own massive operations. For these firms, AI is both a product and a productivity tool, allowing them to scale innovations and efficiencies in a way other companies cannot. Their revenue per employee has jumped from around $550,000 to nearly $800,000, a clear sign of this dual benefit.
The Hurdles Facing Other Firms
For non-tech companies, the path to AI profitability is fraught with challenges. A primary issue is cost. Implementing AI requires significant investment in technology, infrastructure, and specialised talent, which remains scarce. Many businesses are now viewing AI spending less as a growth driver and more as a rising cost centre that is squeezing profit margins without a clear return on investment. Another major obstacle is data readiness. AI models are only as good as the data they are trained on, and many organisations are struggling with 'data debt'—years of fragmented, inconsistent, and siloed information that undermines AI performance. This often leads to pilot projects that work in a controlled test but fail when exposed to the complexity of real-world business data.
A Different Path to Profitability
So,
are other firms doomed to be left behind? Not necessarily,
but their strategy must be different.
Instead of trying to compete with tech giants in building foundational models,
experts suggest focusing on applying AI to solve specific,
high value business problems.
The goal for most companies isn't to invent new AI,
but to become smart adopters of existing technology.
This involves identifying niche use cases where AI can deliver measurable outcomes,
such as automating repetitive tasks,
optimising supply chains,
or personalising customer service.
According to research from McKinsey,
a small group of 'AI high performers' are succeeding not by outspending others,
but by fundamentally redesigning their workflows around AI and maintaining strong leadership oversight to ensure projects deliver real value.
These companies focus on both efficiency and growth,
rather than just cost cutting.
Is It a Winner Take All Market?
The Apollo assessment raises a critical question:
will the AI economy become a winner take all market dominated by a few?
Torsten Slok argues that the true test of AI's economic impact will be decided not by the Magnificent Seven,
but by the S&P 493.
He points out that many core business functions rely on human qualities like judgment,
accountability,
and trust—things AI cannot replicate.
While productivity gains at the individual employee level are being reported,
converting that into company wide profit is the real challenge.
For now,
the gap is widening.
However,
as AI tools become more accessible and commoditised,
the focus may shift from who has the most powerful model to who can apply it most creatively and effectively to solve real world business challenges.
