In a letter to members of the European Parliament, European Central Bank (ECB) President Christine Lagarde reviewed the implications of artificial intelligence (AI) financing for borrowing costs and financial stability. Bond issuance in euros by large technology companies known as hyperscalers has not significantly affected EU government borrowing costs so far. However, expected growth in AI financing creates uncertainty about future bond market effects, while rising debt and ambitious earnings expectations expose the sector to financial shocks. Large bond issues have raised hyperscalers’ own borrowing costs in US dollar and euro markets, but effects on the timing and pricing of comparable corporate issuance have been small and temporary. Euro area investors have absorbed the issuance while also increasing sovereign bond holdings. Sound public finances, credible fiscal strategies and sufficient fiscal buffers would limit potential spillovers to government borrowing costs and help governments absorb increases without abrupt policy adjustments. Completing the savings and investments union and reducing barriers to investment across borders would help channel savings into private and public investment. Firms in the AI supply chain increasingly rely on external financing, including commitments outside their balance sheets to lease or purchase infrastructure. Circular financing through interconnected ownership and mutual commitments, alongside growing private equity and private credit involvement, adds complexity. Revenues falling short of expectations, whether because of cheaper competing models or energy, chip or grid constraints, could put equity valuations and debt sustainability under pressure. Falling equity valuations and market contagion would likely be the main channels for spillovers to the euro area, where entities’ exposure to AI debt remains low.