
Anthropic infrastructure financing reportedly reaches $60B with Broadcom support
The AMW Read
The reported $60B package expands the previously covered $42B financing structure, updating Anthropic's compute economics and connecting infrastructure buildout with supplier credit; the cited IPO disclosures also support the capital-cycle reference.
Anthropic infrastructure financing reportedly reaches $60B with Broadcom support
Broadcom is reportedly supporting a financing package of up to $60 billion for Anthropic's AI infrastructure, according to AI Times, citing Bloomberg. Banks including Bank of America, Citigroup and Morgan Stanley are arranging $42 billion in senior secured loans with residual-value support from Broadcom. Another $18 billion would come from subordinated debt, with Blackstone reportedly contributing $9 billion and other investors supplying the balance. The package is being assembled; the report does not establish that the full amount has closed.
The arrangement ties a frontier model lab's compute access to its hardware supplier's financial support. It extends AMW's October 2 coverage of up to $42 billion in Broadcom financing by describing an additional subordinated layer. Broadcom's reported residual-value commitment supports the senior borrowing against potential equipment depreciation; it should not be read as a blanket guarantee of all $60 billion. For the foundation-model market, the structural signal is that access to infrastructure increasingly depends on financing terms as well as model capability and chip availability.
Investors should distinguish announced capacity commitments from funded equipment purchases and examine who absorbs losses if hardware values fall. The report says Anthropic disclosed potential conflicts because Broadcom is both a hardware supplier and financing partner, with pricing and equipment decisions affecting compute access. Builders evaluating long-term infrastructure agreements face a concrete question: whether supplier-backed financing secures capacity on workable terms while preserving flexibility over equipment, pricing and future compute providers.


