Anthropic’s $11.6 Billion Akamai Cloud Pact Marks a New Phase in AI Infrastructure Spending

Anthropic has agreed to spend $11.6 billion over seven years on Akamai’s cloud infrastructure, a deal announced Thursday that underscores how aggressively major AI companies are locking up compute capacity. The arrangement is Akamai’s biggest ever and could become even larger if Anthropic expands its commitments, while also signaling rising demand for CPUs in AI workloads that are increasingly handled by agents and other general-purpose software.

The pact is notable not only for its scale, but also for its structure. Unlike the more familiar model in which cloud and chip vendors invest directly in the AI labs that buy their services, Akamai is giving Anthropic a warrant tied to future spending. That unusual setup creates a closer financial loop between customer demand and supplier value, and it arrives at a time when AI infrastructure deals are becoming more complex, more strategic, and more expensive.

What Anthropic and Akamai agreed to

The core agreement is simple: Anthropic will commit to paying Akamai $11.6 billion over seven years for cloud services and capacity. Akamai disclosed that the contract depends on the company meeting certain service and delivery conditions, and the deal can be terminated under some circumstances by either side.

The dollar figure is far larger than a previous agreement between the two companies that Bloomberg reported in May, which was valued at $1.8 billion. In less than half a year, the relationship appears to have expanded into one of the largest AI infrastructure commitments announced by a cloud provider.

For Anthropic, the move adds another major layer to its already elaborate web of infrastructure and strategic partnerships. For Akamai, it is a landmark deal that could reshape how investors view the company’s role in AI computing.

Why the deal matters for AI infrastructure

The agreement is part of a broader race among AI companies to secure enough compute to train, deploy and scale advanced systems. As model capabilities improve, the bottleneck has shifted from software ambition to physical capacity: data centers, chips, power, storage, networking and the specialized infrastructure needed to keep AI services responsive and profitable.

What makes this arrangement especially interesting is that it reflects growing interest in CPUs, the general-purpose processors that have often been overshadowed by GPUs in AI discussions. Akamai did not say exactly how Anthropic will use the capacity, but the company suggested the demand is tied to a changing workload profile as AI agents take on more tasks.

That matters because agentic systems often need to do more than generate text. They may browse the web, execute code, coordinate tools, query databases and manage multi-step tasks. Those operations can make CPU-heavy infrastructure more attractive, even as the industry still relies heavily on high-end accelerators for model training and large-scale inference.

How are CPUs becoming more important in AI?

CPUs are becoming more important in AI because not every task in modern AI systems depends on specialized accelerators. While GPUs remain central to training frontier models, many deployment and orchestration tasks are better suited to general-purpose processors, especially when AI agents need to run code, handle logic, manage services or interact with conventional software systems.

That shift does not mean GPUs are less important. It does mean the AI stack is broadening, creating demand for a wider mix of chips, networking and distributed computing resources.

How the Akamai warrant changes the economics

Akamai paired the service contract with a financial instrument that gives Anthropic the right to buy shares at a fixed price. The warrant covers nonvoting preferred stock convertible into up to 7.7 million common shares, or roughly 5% of Akamai’s outstanding stock, at $111.33 per share.

About 2% of that stake is expected to vest once Anthropic makes its first payment under the deal. Additional ownership rights unlock as Anthropic increases spending. For every extra $3 billion of commitments to Akamai cloud services, Anthropic can earn about another 1% of the company, up to a total deal size that could rise by as much as $9 billion.

In practical terms, the structure gives Anthropic a potential equity upside in exchange for customer spending, while also tying Akamai’s financial incentives directly to the size of the relationship.

The warrant arrangement is unusual because it gives the customer a potential ownership stake that grows alongside its cloud spending, rather than the more common pattern in which a supplier invests in the customer.

It is the first time Akamai has attached a warrant to a cloud deal, according to the company’s disclosure and reporting cited by Bloomberg. The design echoes a similar structure AMD used with OpenAI last year, linking warrants to chip-purchase milestones.

What does this mean for Akamai’s business?

The agreement could become a major revenue engine for Akamai, but not immediately. Executives said on Thursday’s investor call that they do not expect any revenue this year from the Anthropic contract.

Instead, Akamai expects the deal to begin generating between $150 million and $300 million in 2027, with activity starting in the second half of that year. By the end of 2028, the company said revenue could be running at an annualized pace of about $1.7 billion.

To support the contract, Akamai plans to invest heavily. The company said it expects to spend about $5.5 billion to build the necessary capacity. It is also increasing its capital spending this year by about $1.7 billion to buy components such as memory in advance, a sign that the deal has already begun influencing its procurement strategy.

Shares of Akamai jumped as much as 17% in after-hours trading after the announcement, according to the Wall Street Journal, reflecting investor enthusiasm for the size and strategic importance of the deal.

Why is Akamai making such a large investment?

Akamai is making such a large investment because the Anthropic contract gives it a long-term demand signal that justifies building new capacity. Cloud infrastructure businesses often need to spend years ahead of revenue in order to secure equipment, expand facilities and stage components before customers begin consuming the service at scale.

The added capital spending also suggests Akamai is trying to position itself as more than a traditional edge and security company. This deal pushes it deeper into the competitive market for AI infrastructure, where scale, reliability and access to hardware are becoming decisive advantages.

Why this deal stands out in the AI financing landscape

AI companies and infrastructure providers have increasingly adopted circular financing structures, but most of those arrangements run in one direction: vendors invest in the startups that buy their chips or cloud services. This deal flips that logic. Instead of Akamai investing in Anthropic, Anthropic gains the potential to own part of Akamai as it spends more with the provider.

That reversal matters because it highlights how bargaining power is evolving in AI infrastructure. Top model developers now have enough scale and perceived strategic value to negotiate terms that look more like partnership finance than simple vendor contracting.

Anthropic is already familiar with these dynamics. Amazon, Google, Microsoft and AMD have all invested in or committed capital to the company while also supplying cloud or chip resources. Anthropic chief executive Dario Amodei said last December that the company does not participate in such arrangements at the same level as some rivals, but the latest Akamai deal shows how deeply these relationships now permeate the industry.

How this fits Anthropic’s broader compute strategy

Anthropic has been on a sustained expansion path as it seeks to secure more computing power for model development and product deployment. The company’s rapid growth has made infrastructure access one of its most important strategic priorities, alongside research, safety work and product development.

This latest deal adds another layer to a long list of capital-intensive commitments across the AI sector. Providers want predictable demand. AI labs want guaranteed access. Both sides often accept complicated financial structures to make the economics work.

That is especially true for a company like Anthropic, whose products are increasingly used in enterprise workflows where uptime, reliability and low-latency access can matter as much as raw model quality. As AI products move from demos to real business systems, the infrastructure underneath them becomes a competitive moat.

Who benefits most from the deal?

Both companies stand to benefit, but in different ways. Anthropic gains a long-term supply agreement that can support scale, while Akamai gains a landmark contract that could materially expand its presence in AI infrastructure and strengthen investor confidence in its growth story.

The equity warrant also means Anthropic could benefit if Akamai’s stock appreciates, while Akamai gets the opportunity to lock in a very large customer over many years.

Key milestones in the Anthropic-Akamai relationship

The deal evolved quickly from a relatively modest cloud arrangement into one of the largest infrastructure commitments in the sector. The timeline below shows how the relationship appears to have accelerated over 2026.

What investors and competitors should watch next

Investors will likely focus on three things: whether Akamai can execute on the infrastructure buildout, whether Anthropic continues to scale at the pace required to justify the contract, and whether the warrant structure becomes a template for future AI infrastructure deals.

Competitors should pay attention as well. If Akamai succeeds, it may encourage other cloud and infrastructure providers to design similar arrangements, especially for customers whose growth can justify large, multi-year commitments. The market could move further away from standard service contracts and toward hybrid agreements that combine customer spend, financing and equity incentives.

For Anthropic, the deal reinforces a larger strategic reality: leading AI developers are no longer just buying compute. They are helping shape the infrastructure economy around them, sometimes as customers, sometimes as partners and increasingly as quasi-financial counterparties.

For Akamai, the announcement is a chance to reposition itself inside one of the fastest-growing segments in technology. And for the broader AI industry, it is another reminder that the race to build better models is inseparable from the race to secure enough physical capacity to run them.