As artificial intelligence development costs soar into the trillions, major tech players are racing to find revenue streams that can justify their staggering investments. Meta and xAI have both made notable moves in recent weeks, striking deals to rent out data center capacity and charge developers for access to advanced models — a sign that the era of unchecked AI spending may be hitting a wall.
Businesses and consumers remain unconvinced
Despite the industry’s enthusiasm, the real-world impact of AI on businesses remains underwhelming. A National Bureau of Economic Research study of nearly 6,000 executives found that nine in ten reported no measurable effect on employment or productivity over the past three years. Uber’s operations chief has publicly noted the difficulty of justifying AI costs, saying that higher token usage has not translated into genuinely useful consumer features.
Meta’s AI Superintelligence Chief Alexandr Wang acknowledged the disconnect, admitting that AI has yet to demonstrate how it truly empowers ordinary users. While developers have embraced AI as a transformative productivity tool, that enthusiasm has not trickled down to the average consumer or business owner.
Rising competition adds pressure
Adding to the strain, Chinese AI developers are releasing competitive models at a fraction of the cost, threatening to undercut the massive U.S.-led AI infrastructure push. With consumer sentiment declining and efficiency gains failing to materialise at scale, the AI industry faces growing pressure to deliver tangible value — or risk a significant market