Anthropic to pay Akamai $11.6 billion over seven years

AI company Anthropic has signed a seven-year, $11.6 billion agreement to use Akamai's cloud and processor infrastructure. As reported by TechCrunch, it is the largest deal in Akamai's history. The commitment is not unconditional: it depends on Akamai meeting certain delivery and service-duration terms, and either side can terminate the agreement under certain conditions.

How striking is the size of the deal?

The figure is more than six times the initial $1.8 billion deal reported by Bloomberg in May. Akamai will invest to expand its infrastructure under the agreement, and said it will also increase this year's capital expenditure in order to buy components such as memory in advance.

Why CPUs rather than GPUs?

One notable aspect of the deal is its focus on CPUs (general-purpose processors), an area that features less prominently in AI infrastructure investment. Akamai did not say exactly what Anthropic will use the capacity for; however, demand for CPUs is known to be rising as AI agents take on tasks such as running code and browsing the web.

When will Akamai start earning from the deal?

Akamai will not earn revenue from the deal this year; executives told an investor call that they expect revenue to reach an annual run rate of $1.7 billion by the end of 2028. The agreement also grants Anthropic a warrant for Akamai stock, convertible into 7.7 million common shares, equivalent to up to around 5% of the company's outstanding shares. Part of this right vests when Anthropic makes its first payment, and the rest unlocks gradually as Anthropic increases its spending. Following the announcement, Akamai shares rose as much as 17% in after-hours trading.

Is this a new pattern for AI deals?

Usually chipmakers and cloud providers invest directly in the AI labs that buy their products; here, the supplier Akamai is giving its customer Anthropic an equity stake that grows as the customer grows, a method similar to the structure AMD previously set up with OpenAI. Anthropic also has similar investment arrangements tied to chip or cloud capacity with Amazon, Google, Microsoft and AMD; CEO Dario Amodei had said the company does not take part in such deals on as large a scale as some other companies.

What does it mean for companies?

For UNIT Journal readers, the key message is not the size of the figure but the structure itself. Large AI labs now cover infrastructure costs with equity and long-term commitments rather than cash, a model designed to help them keep service prices low in the short term. For companies that are growing on the back of AI tools or cloud services, the lessons are:

• Keep track of how your AI provider finances its infrastructure costs; cross-investment deals like this can affect the provider's pricing and service priorities in the long term.
• When signing multi-year cloud or AI contracts, use exit clauses as a model for your own contracts, such as the right to terminate if the other party cannot meet its delivery terms.
• Rising CPU demand shows that the spread of AI agents increases not only model costs but also the need for general computing power running in the background; factor this into budget planning.