Anthropic has signed a seven-year, $11.6 billion cloud agreement with Akamai Technologies — a contract more than six times larger than the one Akamai booked earlier this year, and one that hands the AI lab an ownership stake in its supplier.
Akamai announced the deal after US markets closed on Thursday, September 24. The commitment covers distributed cloud capacity that Anthropic says it needs for its accelerating CPU workload requirements — the layer that serves inference, data processing and agent traffic rather than large-scale model training. According to Reuters, the total could rise to as much as $20 billion, and Akamai issued Anthropic a warrant for up to about 5 percent of its shares.
The market reaction was swift. Akamai stock spiked in Friday pre-market trading, with reports of gains around 20 percent before the session opened; in Reuters' closing market wrap, the shares finished the day up 3.2 percent as AI names broadly rallied.
The contrast with earlier commitments is stark. In May, Bloomberg reported a $1.8 billion, seven-year cloud agreement between Akamai and an unnamed major AI company widely believed to be Anthropic. The new contract replaces that figure with one several times larger — a sign of how aggressively AI labs are locking in compute while capacity remains tight.
For Anthropic, the CPU focus complements the GPU-centric deals that dominate AI infrastructure headlines: whatever runs on frontier accelerators still has to be served, logged and processed somewhere, and edge-distributed capacity is often the cheapest place to do it.
For Akamai, the agreement is a repositioning. A company long known for content delivery and edge security is now a contracted AI cloud supplier with a marquee customer — and, through the warrant, an equity link that ties the supplier's fortunes and the customer's upside partly together.
