Tether CEO Warns of Four Cracks in Big Tech’s AI Boom

Tether CEO Paolo Ardoino has warned that Big Tech’s artificial intelligence spending boom may be built on weak economics, as concerns over an AI market bubble spread across global markets.

In a July 4 post on X, Ardoino said the AI infrastructure race contains four major “structural mismatches”. These are gaps between costs, revenues, investment timelines, and competition.

AI big tech subsidizes compute to increase user count building expensive infrastructure / capex subject to fast decay (3/5 years).– Token price mismatch.– Profitability timeline mismatch.– Cost of capital maturity mismatch.– Open-source AI taking growing chunks of revenues.…

— Paolo Ardoino 🤖 (@paoloardoino) July 4, 2026

His warning comes as the world’s largest technology companies pour hundreds of billions of dollars into data centres, chips, and power capacity. The central question for investors is whether AI can generate enough revenue to justify that spending.

Businesses are Paying Too Less for AI

Ardoino argued that companies are charging too little for AI computing compared with the real cost of providing it. In simple terms, some AI services may look cheap because companies are subsidising usage to win customers.

That makes growth look stronger than the underlying business model. If companies later raise prices, users may spend less. If they keep prices low, margins may remain under pressure.

AI Profits Could Take Longer to Realize

Big Tech companies are spending heavily now, while the profits from AI may take much longer to arrive. Data centres, GPUs, and power contracts require huge upfront investment.

This creates a gap between capital spending today and commercial returns in the future. The bigger that gap becomes, the more pressure companies face to prove that AI can become a durable source of revenue.

The dot com bubble and the AI trade look similar.The difference is that this time the earnings are showing up, compressing valuations as investors disagree on just how impactful AI will be in our lives and work.Here are 5 dirt cheap AI stocks you can buy in 2026.1. Nvidia -… pic.twitter.com/6RB8mouPyh

— Michael Sikand (@michaelsikand) July 2, 2026

AI is Outdating Itself Fast

AI chips can become outdated within 3 to 5 years. Yet the debt and equity used to finance AI infrastructure often assume a much longer payback period.

That matters because companies may need to replace expensive hardware before it has fully paid for itself. If demand slows or pricing falls, the economics become harder to defend.

The Industry Faces Intense Competition

Open-source AI models are improving quickly and could weaken the pricing power of commercial AI providers. If cheaper or free alternatives become good enough, customers may resist paying premium prices.

That would make it harder for companies to recover the money they are spending on infrastructure. It could also reduce the revenue expectations that have supported high AI valuations.

Ardoino’s warning is part of a wider debate now moving through markets.

Chinese hedge funds including Wealspring Asset and Shanghai Banxia Investment Management Center have warned that AI stocks may be in bubble territory. Wealspring reportedly called global AI stocks a “super bubble”, while Banxia said a possible trigger for a correction may already have appeared.

The concern is simple. AI has become a major driver of stock market performance, especially for large technology companies. If investors begin to doubt the return on AI spending, the impact could spread beyond the tech sector.

🚨 The real warning signal for AI may not come from the tech giants but from their customers. ➡️ For now, the infrastructure sellers continue to capture most of the value but the sectors expected to use AI heavily are not yet showing a spectacular acceleration in earnings.… pic.twitter.com/wnFq7bH729

— Christophe Barraud🛢🐳 (@C_Barraud) July 4, 2026

AI Spending Could Hit Trillions

JPMorgan has projected that global AI-related spending could reach $5.5 trillion by 2030. At the same time, Alphabet, Amazon, Meta, and Microsoft are expected to spend up to $720 billion this year.

That level of spending gives AI a central role in corporate earnings, energy demand, chip demand, and credit markets.

The Bank of England warned in October 2025 that AI-related valuations had moved close to levels seen during the dot-com bubble. It also said AI infrastructure may require trillions of dollars, with a meaningful share financed by debt.

Some investors take a less negative view. They argue that today’s AI trade differs from the dot-com era because the largest companies funding the boom already have strong earnings and established businesses.

Morgan Stanley has also estimated that nearly $3 trillion in AI infrastructure investment could move through the economy by 2028.

Still, Ardoino’s point is that the risk sits inside the economics of AI infrastructure itself. If pricing, profits, hardware lifespans, and competition do not line up, the market may be underestimating how hard it will be to turn AI demand into lasting returns.