Fed's Bowman warns AI stocks could correct sharply if AI revenues disappoint, with AI debt set to top $800 billion
In a speech at the Council on Foreign Relations, the Fed's Vice Chair for Supervision said AI revenues are not yet enough to cover the cost of models and data centres, and that off-balance-sheet AI obligations are estimated at more than three times the debt.
Why it matters
It is one of the most detailed warnings yet from a senior Fed official about how the AI buildout is financed. Bowman put AI-related debt on track to exceed $800 billion this year and said leases, contracts and residual value guarantees could be worth more than three times that. She also said AI-related demand accounts for roughly a quarter of above-target core inflation. Watch whether the Fed's stress tests and data collection on AI counterparties lead to new supervisory guidance for banks.
Key facts
- Speech at the Council on Foreign Relations, Washington, 10 November 2026
- AI-related debt on track to exceed $800 billion this year
- Lease, contract and residual value obligations estimated at more than three times that debt
- Tariffs, energy and AI each about one-fourth of above-target core PCE inflation
- Core PCE inflation likely 3.3% in October
- About 25% of businesses and 10% of workers use AI every day