Cloud Giants Hold the Real Power as AI Costs Rise and Growth Slows

Silicon Valley's belief that building the smartest artificial intelligence system guarantees industry dominance is facing a reality check. The cost of building AI models is going up while the price of using them goes down.

The Growth Slowdown

Anthropic is heading toward a public offering in November and aims for a $2 trillion valuation. However, the company recently told investors that its third-quarter revenue reached $16.8 billion, marking a quarterly growth rate of around 50%. That figure is a sharp drop from the 140% growth seen earlier this year. A chart published by Citadel Securities in August highlights a painful economic trap for firms like OpenAI and Anthropic, where computing power is getting more expensive even as token prices fall.

Shifting Market Demands

Businesses are increasingly choosing AI models that are just good enough instead of paying for the absolute best. Cheaper, open-weight models now account for more than a third of enterprise AI token usage, rising from 23% a year ago according to an Enterprise Technology Research survey of 200 companies in September. Respondents expect that share to hit 41% over the next 12 months. Over the past year, major Anthropic customers Microsoft and Meta Platforms cut back on internal Claude use to favor their own proprietary solutions and lower token costs.

Power of Cloud Giants

Real power now rests with the companies that distribute and power the technology, specifically cloud giants like Microsoft, Alphabet, and Amazon. Shares for this Big Tech trio have risen an average of 10% since the start of 2026. They control the infrastructure, customer relationships, and distribution channels, allowing them to profit whether a business uses Claude, an older ChatGPT version, or a cheaper model from China. OpenAI and Anthropic face a massive gap between their revenue and a collective $1.3 trillion they are expected to spend on data centers, chips, and computing power through the end of 2030.

Context

Major technology shifts throughout history often see infrastructure providers emerge as more resilient financial winners than the firms building the frontline consumer products. Companies with diversified revenue streams and established enterprise relationships are historically better positioned to weather market corrections. Cloud giants maintain profitable operations well outside of artificial intelligence, giving them a strong financial cushion that standalone AI startups lack.