Morgan Stanley expects that global debt issuance linked to AI companies will exceed $500 billion in 2026, driven by aggressive capital expenditure plans from major hyperscalers such as Amazon, Google, and Meta. The bank projects this issuance could rise to around $570 billion as AI firms increasingly turn to alternative funding sources to support expansion.

Recent large-scale financing efforts underline this trend. Alphabet, the parent of Google, recently launched an $85 billion fundraising plan aimed at expanding AI infrastructure such as data centres and computing facilities, and earlier issued a rare 100-year bond to support its AI investments. According to estimates cited by Reuters, AI-related global debt issuance had already reached nearly $236 billion by the end of May 2026, marking a fourfold increase from the previous year.

Strong demand for advanced AI models and the rapid development of agentic AI systems have also prompted companies like Anthropic and OpenAI to scale up spending significantly, with both reportedly preparing for potential public market listings. Their valuations are estimated at about $965 billion and $852 billion respectively.

The heavy computing requirements needed to train and deploy these models are pushing hyperscalers to expand data centre capacity and increase capital spending, with Morgan Stanley forecasting hyperscaler capex could exceed $1 trillion by 2027. It also notes that these firms are diversifying funding sources, including issuing more non-US dollar debt. Amazon recently raised C$14 billion through Canadian dollar-denominated notes and also secured €14.5 billion in one of the largest euro corporate bond deals, highlighting how hyperscalers are tapping global debt markets to finance large-scale AI infrastructure expansion.