AI industry revenue has reached a critical threshold, according to a report released by research firm Exponential View, which said the commercial model behind the hundreds of billions of dollars that technology companies have spent on AI infrastructure is starting to see early validation.
The report said AI-related sales revenue generated by global hyperscale cloud providers and emerging cloud providers has risen to about $25 billion. That figure has now exceeded the estimated depreciation cost tied to AI data center and chip investment for a second consecutive quarter, with that cost put at roughly $21 billion.
Exponential View said the milestone shows that revenue generated by the AI industry is beginning to cover the cost pressure created by infrastructure capital spending. In that reading, the AI economy is moving out of a phase defined mainly by capex-led expansion and into one where revenue is beginning to validate the spending.
According to the report, current AI revenue is concentrated in AI cloud services, GPU compute rentals, large-model APIs, enterprise AI software, and generative AI applications.
The firm also said the pace of AI commercialization is picking up as enterprise customers continue to increase spending on AI. Even so, the report said the industry is still far from a high-profit stage. Costs tied to GPUs, data centers, electricity, and model development remain high, leaving limited room for profit.
Exponential View said current revenue should be seen more as proof that infrastructure investment can be sustainable, rather than evidence that the sector has already reached large-scale profitability. The report said the next phase of competition in AI will shift from the question of whether real demand exists to which companies can achieve profitability at scale in a highly competitive market.
As model capabilities improve and costs decline, AI service prices could fall further, meaning companies will need more efficient use cases and business models to lift margins.