Paolo Ardoino, the CEO of Tether, has expressed apprehensions over the long-term stability of the AI business framework of Big Tech. Paolo Ardoino argues that the firms are aggressively funding compute resources while spending billions on sprawling diverse data centers, electricity, and GPUs to advance consumer growth. According to Paolo Ardoino’s exclusive X statement, by providing synthetically low prices of tokens, often via fiat-rate or even discounted AI access, the respective entities attempt to increase adoption. Nonetheless, he has warned that the respective strategy leads to a dangerous imbalance as the token revenues produced from consumers do not align with the swiftly elevating infrastructure costs.
AI big tech subsidizes compute to increase user count building expensive infrastructure / capex subject to fast decay (3/5 years).
— Paolo Ardoino ? (@paoloardoino) July 4, 2026
– Token price mismatch.
– Profitability timeline mismatch.
– Cost of capital maturity mismatch.
– Open-source AI taking growing chunks of revenues.…
Recently, Paolo Ardoino has intensified his criticism of Big Tech AI companies. Last week, he claimed big tech AI is waging war on open-source AI.
AI Token Rates vs Actual Compute Costs Raise Possibility of Mismatch
Paolo Ardoino points out that the mismatch between the real expenses and token pricing could eventually undermine profitability while also destabilizing the market in the medium term. At the core of this mismatch is the billing units dealing with the AI framework usage. In this respect, the tokens that the systems like Claude or GPT are processing are priced notably below the real compute cost.
He added that the infrastructure investments diminish rapidly, often within 3 to 5 years, indicating the relentlessness of the capital expenditure cycle. Companies must consistently reinvest in exclusive facilities and hardware, yet the synthetically low token prices fail to create enough cash streams to cover the respective obligations.
Another key dimension that Ardoino highlights is the mismatch dealing with the cost of the overall capital maturity. Specifically, Big Tech entities fund AI expansion via equity and debt, but the respective financial instruments’ maturity does not align with the compute infrastructure’s short lifespan. As GPUs and servers depreciate swiftly, repayment burden grows heavier.
AI Profitability Crisis Risk Looms
According to Paolo Ardoino, such a structural imbalance could compel companies into hard choices, including raising token rates sharply, continuing burning capital, or decreasing subsidies. Ardoino’s warning signifies a wider reckoning for the rapidly expanding AI market. Though subsidized token rates drive adoption, the fundamental economics disclose cracks that face risk of expansion over time. So, unless companies recalibrate their existing frameworks to balance stable infrastructure financing, the sector could go through a profitability crisis.
