Tether CEO Ardoino Warns AI Spending Mismatches Could Hit Crypto
Tether CEO Paolo Ardoino is worried about the AI spending boom, and he is not dressing it up. Big Tech may be building too much AI infrastructure before anyone has proved the money comes back. My take: crypto investors should care more than they probably want to. If AI stocks crack, the selloff probably will not stay politely fenced inside tech. Bitcoin, ETH, and smaller tokens could all get dragged into the same risk off trade.
Ardoino made the case on X today. He said AI data center spending sits on four mismatches: subsidized compute, expensive chips, uncertain revenue, and hardware that may only stay useful for three to five years. Clean. Uncomfortable. Why does this matter? Because hyperscalers are spending at record levels while investors still do not have clean evidence of returns.
The four problems he named were direct: token prices do not match compute costs, profit timelines do not match investment timelines, capital maturities do not match asset life, and open source AI could push revenue lower. Most AI bulls frame this as a temporary scaling problem. That is only half right. These sound less like small accounting complaints and more like balance sheet stress points that stay quiet while money is easy, then suddenly become the entire story.
The spending numbers are enormous. JPMorgan’s midyear outlook from June 24 raised its estimate for global AI-related capital spending through 2030 to $5.5 trillion, up from $5.1 trillion. It expects AI-related debt financing to reach $4.1 trillion. Hyperscaler capital spending could hit $650 billion this year and top $1.1 trillion in 2027. Microsoft alone plans to spend about $190 billion in 2026, a 61% increase from the previous year. Goldman Sachs estimates Meta, Microsoft, Amazon, and Alphabet will spend $5.3 trillion on capital expenses from 2025 to 2030, with $725 billion planned this year. Last year, that figure was $410 billion. Alphabet also raised $84.75 billion for AI infrastructure, reportedly the largest US equity capital raise on record.
The profit question is where this starts to bite. Ardoino’s warning lands while companies are spending heavily on AI and often still cannot show a clean return. The average company is expected to spend $11.5 million on AI this year. Many still cannot show whether that spending pays for itself. Bureau of Economic Analysis data shows Information sector growth slowed to 1.5% in Q1 2026, down from 3.2% in Q3 2025. That is not just macro trivia. Crypto usually hates a cooling growth story. If stocks weaken hard, Bitcoin’s “digital gold” pitch gets tested quickly, and a sharp move could push BTC back below the $60,000 level it has struggled to hold.
Ardoino also warned that open source AI could squeeze revenue, and that pressure is already showing up inside companies. Amazon dropped its internal leaderboard for employee AI use. Uber burned through its 2026 AI coding budget in four months and then set a $1,500 monthly cap per employee. Meta warned 6,000 staff about rising costs. IDC expects 70% of major AI adopters to use multiple models by 2028, which could start a price war. Is this overkill? For a few experimental chatbots, maybe. For the infrastructure spending now sitting behind Amazon, Uber, Meta, and the broader cloud stack, no. If AI teams start cutting costs, the effect could spread into venture funding for early crypto projects and push big company blockchain experiments down the priority list.
Regulators are uneasy too. The Bank for International Settlements warned in its annual report that a sharp drop in AI investment could hit global stock markets harder than past recessions. It listed AI as one of the three main economic risks. Zhang Tao, the BIS chief representative for Asia and the Pacific, said “the speed of a correction could be much faster than previous banking crisis episodes.” I’ll be honest: crypto traders should not shrug that off. A fast unwind in AI could force funds to cut exposure across risky assets at the same time. ETH and smaller altcoins would probably take more damage than BTC.
Not everyone accepts the bear case. Wedbush analyst Dan Ives has called the AI buildout an “arms race” that major companies cannot afford to skip, and he expects profits to show up within six to twelve months. JPMorgan also expects strong profits, with operating cash flow above $900 billion by 2027. Counter to the usual doom read, those numbers are not trivial. Great Hill Capital chair Thomas Hayes takes a more cautious middle view. He thinks one or more big companies may announce lower capital spending in upcoming earnings reports. That is the next test. If a major spender cuts guidance, Ardoino’s warning becomes much harder to dismiss.
What this means
Ardoino is warning that AI infrastructure spending may be running ahead of the business case. If that bubble pops, crypto will probably feel it. Yes, this contradicts the clean “Bitcoin is separate from tech” story. Bear with me. Bitcoin has survived plenty of market stress, but AI is now tied into stocks, debt markets, chips, cloud spending, and corporate budgets. A fast correction there could pressure BTC near $60,000 and hit higher beta tokens even harder.
The next earnings season matters. Watch Microsoft, Alphabet, Amazon, and Meta for any sign that capital spending is slowing. Hayes said that kind of pullback may be coming, and it would be the clearest sign that the market is starting to question the AI buildout. Also watch BEA Information sector growth and any new BIS warnings. Skip the victory laps. If those signals keep weakening, crypto investors should expect more volatility, especially in ETH, DeFi tokens, and smaller altcoins that depend on easy capital.