AI borrowing boom intensifies competition for European investment funds

US technology companies financing the artificial intelligence expansion are drawing on investment funds also sought by European governments and businesses. With more than $1 trillion, or €900 billion, flowing into the AI build-out, the industry's growing demand for debt financing could put further pressure on borrowers already facing higher interest costs.

Amazon, Alphabet, Meta and Oracle have issued about $250 billion in dollar-denominated bonds since the start of 2025, according to AI Industry Reviews, an independent AI tools evaluation platform that compiled US regulatory filings. The four companies also sold €37 billion, or $41 billion, in euro-denominated debt. Microsoft recently followed with a $25 billion bond sale entirely in dollars. Goldman Sachs forecasts $420 billion in new AI hyperscaler bond issuance next year, while ING expects euro-denominated sales of about €60 billion.

The borrowing surge comes as sovereign bond yields in France, Germany and Italy have risen sharply in recent months. France, where public debt stands at 119% of gross domestic product and the deficit remains above 5%, has seen its 10-year bond yield increase from around 3.7% to almost 5% over six months. Its annual interest bill is projected to grow from €79 billion currently to €91 billion next year and exceed €100 billion in 2028, increasing the need for spending cuts, tax increases or further borrowing.

European Central Bank economists said in August that the euro bond market had absorbed the arrival of AI companies well. However, they cautioned that the scale and concentration of their borrowing could restrict access to financing for other businesses and sectors. Their warning also extended to government debt and international borrowers, including the European Union and the European Investment Bank, which could face higher financing costs.

Viktor Hjort, global head of credit strategy at BNP Paribas, described the situation to DW as “bond abundance.” He said hyperscalers already had a sufficiently large presence in the euro market to compete with companies and governments for funds that mainly invest in bonds. Yields rose again in October as energy prices rebounded in connection with the Iran war and France faced a political crisis. Hjort said the simultaneous rise in corporate and government borrowing costs exposed that competition. During uncertainty, investors would normally seek government bonds, lowering their yields.

European businesses are already feeling the effects of more expensive debt. Goldman Sachs said last week that interest expenses as a proportion of sales among companies in the Stoxx 600 index had increased by nearly half since 2022, reducing earnings by more than 5%. Additional competition from AI borrowers could push firms to reduce spending on plants, equipment and research, along with shareholder payouts, to prioritize repayments.

Cities and regions could also face consequences. An OECD study published in January examined local bond markets in Belgium, Germany, Spain, Sweden and Canada. It found that municipal and regional borrowing costs tended to rise in places where AI adoption was advancing fastest, measured through the proportion of AI-related job vacancies.

Study co-author Christos Makridis of Arizona State University's W. P. Carey School of Business told DW that this pattern in several European countries contrasted with his US research, which linked AI investment to lower municipal borrowing costs. AI borrowing is only one contributor to Europe's fiscal pressures, but the additional bond sales planned for next year could increase the burden on public finances.