Anthropic signs $11.6 billion seven-year cloud deal with Akamai

By Splitfeed EditorialLast updated Sep 28, 2026, 11:20 UTC

On September 24, 2026, Anthropic signed a seven-year, $11.6 billion cloud services agreement with Akamai Technologies. As part of the arrangement, Akamai issued Anthropic a warrant for up to 5 percent of its common stock. Akamai simultaneously announced a supply agreement with Lenovo and a separate $1.7 billion build contract with Jabil. Akamai shares rose between 16 and 20 percent in after-hours trading following the announcement. The deal brings Anthropic's cumulative compute spending to roughly $517 billion across just 11 months of agreements, and it moves a meaningful share of that spending away from Amazon, which holds its own major cloud partnership with the company.

The scale and pace of Anthropic's infrastructure commitments give this deal significance beyond a single contract. CEO Dario Amodei has previously warned that Anthropic could face bankruptcy if its revenue projections fall short, which means the company is making multi-billion-dollar, multi-year bets while publicly acknowledging the fragility of its financial position. The warrant structure also ties Akamai's equity future more directly to Anthropic's fortunes than a straightforward services contract would. At the same time, critics have raised questions about whether AI data center economics can justify commitments of this size across the industry, and whether the pace of expansion is consistent with any serious concern about AI risk. These are the fault lines the deal reopens.

The Pro-AI camp reads the agreement as straightforward validation. The account robot2trade called it a huge validation of Akamai's AI infrastructure push, reflecting a broader view that large, long-term contracts signal maturing commercial demand for AI compute. The Anti-AI camp is more fractured in its criticism but consistent in its skepticism. The accounts coolzonemedia and yunod both pointed to economist Paul Kedrosky's argument that AI spending is distorting bond markets and that data center economics do not hold up under scrutiny. The account alextech-ai noted that Amodei signed an AI slowdown appeal before Anthropic shipped Opus 5.5 and Anthropic continued racing toward an IPO, a tension others in that camp echo. The account codehappy questioned whether Anthropic's stated mission to contain an intelligence explosion is coherent given its expansion pace. The Middle Ground camp has posted no reactions to this story; that camp would typically look for evidence of whether revenue is growing fast enough to justify the spending commitments before drawing conclusions about sustainability.

The clearest near-term signal to watch is Anthropic's revenue trajectory. If the company approaches or reaches the projections underpinning these contracts, the sustainability criticism loses force; if it falls short, Amodei's own bankruptcy warning becomes the story rather than the deal.