Anthropic eyes $518 billion AI buildout; backed by 80% non-cancellable deals

Anthropic expects to spend at least $518 billion over the next decade to secure AI computing infrastructure from six partners, with about 80% of the commitments either non-cancellable or payable regardless of usage, according to a Reuters report.

The AI startup said it plans to spend at least $111.1 billion with Google , $110 billion with Amazon and $31.4 billion with Microsoft under long-term infrastructure service obligations over the next seven to 10 years "regardless of ⁠usage," as per the report.

"Anthropic said it expects to spend at least $518 billion over a decade building AI infrastructure with ​six partners, with about 80% of that sum non-cancelable or requires payment regardless of usage," it said.


The company is telling investors the commitments are necessary because access to computing power is becoming the key constraint on AI development, as future demand for advanced AI systems is likely to exceed available supply and will be "limited principally by the availability of compute."

"Anthropic is separately carrying about $161.2 billion of Broadcom-related equipment lease obligations that are largely non-cancelable," according to Reuters.

Anthropic confidentially filed for an initial public offering (IPO) with the US Securities and Exchange Commission (SEC) in June, although the filing has not been made public.

The company has also committed to make payments to Google between April 2026 and July 2033, and to Amazon between May 2026 and April 2036.

"If our actual spend falls short, we must pay Google the difference," it said, adding that similar terms apply to its Amazon agreement, Reuters reported.

The company also disclosed agreements with Elon Musk's xAI that could lead to spending of up to $84.5 billion on Nvidia-based computing capacity through 2029. However, most of these commitments can be cancelled with 90 days' notice, the report said.

Apart from this, the filing also highlighted Anthropic's expanding partnership with Advanced Micro Devices (AMD), which has committed to purchasing up to $5 billion of Anthropic shares and supplying AI computing capacity expected to exceed $20 billion.

The company is increasingly building its own AI ⁠infrastructure, shifting from a cloud-only model toward dedicated data centers and directly leased chips.

It flagged a key risk in its dependence on Amazon, Google and Microsoft, which provide computing infrastructure and distribution while also developing rival ​AI models.

"If the compute we have access to from third parties is curtailed, repriced, or terminated ... our business, financial condition, and results of operations could be adversely affected," it said, as reported by Reuters.