Anthropic’s seven-year commitment to Akamai signals that the next phase of AI infrastructure may depend on securing broad, reliable computing capacity, not just buying more GPUs. The agreement also shows how suppliers and AI companies are beginning to share financial risk as demand for agentic workloads remains difficult to forecast.
The contract is significant because it expands the infrastructure relationship between an AI developer and a provider better known for distributed cloud services than for operating one of the industry’s dominant hyperscale platforms. Akamai said in its official announcement that Anthropic has committed to spend $11.6 billion over seven years on its cloud infrastructure. Akamai described the arrangement as the largest contract in its history.
The agreement is not unconditional. Akamai’s filing with the US Securities and Exchange Commission says the project depends on the provider meeting delivery and service availability requirements. The agreement also gives either company termination rights under specified circumstances, making the headline value a potential commitment rather than guaranteed revenue.
That distinction matters. AI companies are racing to secure capacity years before they can be certain how quickly demand will grow, while infrastructure providers must spend heavily before the related revenue arrives. The contract therefore represents both a bet on AI expansion and a test of whether suppliers can build enough capacity without creating excessive financial exposure.
A long wait for revenue
Akamai expects to spend approximately $5.5 billion developing the capacity required by the agreement. The company is also adding approximately $1.7 billion to its 2026 capital spending plan to purchase components, including memory, in advance.
The investment will not produce revenue for Akamai this year. Company projections call for between $150 million and $300 million in revenue during 2027, beginning in the second half of the year. The annualized pace is expected to reach approximately $1.7 billion by the end of 2028.
Akamai’s investor presentation details the revenue and capital spending ramp, illustrating the mismatch between the timing of costs and sales. Akamai must commit capital well ahead of the point at which the contract becomes a material contributor to its business.
That pattern has become familiar across AI infrastructure. Chipmakers, data center operators and cloud providers are spending aggressively on equipment and facilities, while AI companies sign large capacity agreements to avoid shortages. The risk is that demand projections may change before those assets are fully utilized. The potential reward is a multibillion dollar stream of contracted business if AI usage continues to expand.
The deal includes a further growth option. If Anthropic commits another $9 billion, the total value could rise to approximately $20 billion. The option gives Anthropic room to expand if its usage grows, while giving Akamai a path to a much larger relationship without requiring the full amount immediately.
A customer receives an ownership stake
The financial structure is unusual because Akamai is giving its customer a potential ownership interest. Under the agreement, Akamai granted Anthropic a warrant for nonvoting preferred stock that can convert into as many as 7.7 million common shares. The shares could represent as much as approximately 5% of Akamai, with a strike price of $111.33 per share.
About 2% is expected to vest after Anthropic makes its first payment. Additional portions depend on increased spending. The warrant therefore links Anthropic’s potential equity value to the scale of its infrastructure commitment.
As the SEC filing sets out the legally material terms, the arrangement includes conditions covering delivery, availability, termination, vesting and conversion. Those details are important because they determine how much of the announced value becomes binding and how much of the potential ownership stake Anthropic can ultimately receive.
The structure reverses a common pattern in the AI economy. In many major deals, chipmakers and cloud companies invest in the AI laboratories that purchase their hardware and services. Here, the infrastructure provider is giving the customer a potential stake that grows as the customer spends more.
That approach could align both sides. Anthropic receives a possible financial benefit from helping Akamai scale the relationship, while Akamai gains a customer with an incentive to expand its commitments. At the same time, the warrant may dilute existing shareholders if the milestones are met and the shares are issued.
Why CPUs matter
The agreement is also notable because the capacity is associated with CPUs as well as broader cloud infrastructure. GPUs remain central to training and running many advanced AI models, but CPUs handle general purpose computing tasks. Those include running code, managing data and browsing the web.
Those functions may become more important as AI agents perform longer sequences of actions. An agent that researches a topic, opens websites, writes and executes code, calls software tools and checks the result needs more than a model accelerator. It needs a collection of computing resources that can support unpredictable, interactive work.
Akamai has not specified exactly how Anthropic will use the capacity. The company’s announcement therefore does not establish that the infrastructure will replace GPU focused systems or that CPUs will become the dominant resource. It does suggest that the commercial market is preparing for workloads with a wider mix of requirements.
TechCrunch reported on the agreement, highlighting the scale of the commitment and its implications for AI infrastructure beyond the largest hyperscalers. The broader lesson is that frontier AI companies may increasingly assemble capacity from several providers rather than rely on a single cloud platform.
A test of the AI buildout
The Akamai agreement does not prove that agentic computing will generate enough demand to justify every planned expansion. Its conditional terms and delayed revenue schedule underline the uncertainty. Akamai must spend billions before the full commercial return is visible, while Anthropic is committing to capacity whose value will depend on future model adoption and usage patterns.
Still, the contract points to a changing infrastructure market. The next constraint may not be access to one type of processor. It may be the ability to combine accelerators, CPUs, memory, networking and cloud services at sufficient scale and predictable cost.
That makes the deal more than a large procurement contract. It is a financial and technological experiment in building an AI supply chain around uncertain demand. If agentic workloads grow as expected, Akamai could become an important capacity provider outside the traditional hyperscaler group. If growth slows, the same commitment could leave both companies managing expensive infrastructure built for a future that arrives more slowly than planned.
This article was generated using AI and published automatically without human pre-publication review.
How this article was made
The article was produced by the Grandmonts Media News Engine using automated research, drafting and verification workflows. No human editor reviewed the article before publication. Grandmonts Media remains responsible for the published content. Errors can be reported at office@grandmonts.cz.