Economist Warns AI Boom May Face External Triggers and Debt Risks
Dario Perkins of Global Data.TS Lombard cautions that the AI sector's growth could be undermined by external shocks, rising debt outside major tech firms, and potential political backlash.
Investors are being urged to look beyond the rapid expansion of artificial-intelligence spending, as economist Dario Perkins highlighted possible vulnerabilities that could disrupt the market. Perkins, managing director of global macro at Global Data.TS Lombard, said the sector’s revenue growth is still largely driven by internal capex recycling rather than payments from businesses and consumers for AI services.
He noted that debt across parts of the AI ecosystem could reach $2 trillion by the end of 2027, with smaller cloud providers, model developers and infrastructure suppliers taking on increasing leverage. While he does not view AI-related debt as a systemic financial risk, earnings shortfalls or order cancellations could affect these weaker firms, especially if cash-rich hyperscalers reduce data-center leases.
Perkins also warned of a potential chip oversupply, citing reports that companies have over-ordered semiconductors to guard against shortages, which could create excess inventory if demand slows. He added that higher interest rates might further strain companies that rely on borrowing for expansion.
Beyond corporate factors, Perkins highlighted political risk, stating that public sentiment toward AI is uneasy and could translate into opposition if a high-profile incident occurs.
"The trigger doesn't have to be endogenous to the boom. It could be external," Perkins wrote in a note on Monday. "The question is whether those revenues are coming from businesses and end-users outside the ecosystem — presumably to pay for productivity-enhancing tools — or from simple capex recycling," he added. "Let's not forget we are talking about tech hardware and chips, notoriously cyclical, and prone to bullwhip effects," Perkins said. "AI boosters on social media live in a bubble. They don't realize that ordinary people don't like AI. They don't find it useful, they REALLY don't want to live near datacenters, and they are worried about their jobs. And they vote," he wrote. "One high-profile accident turns AI toxic," Perkins concluded.