Akamai spent decades building one of the world's largest distributed internet networks.
Now artificial intelligence may be turning that infrastructure into a much bigger business.
The company announced a massive seven-year agreement with Anthropic worth $11.6 billion in contractual commitments, sending Akamai shares soaring more than 20% in after-hours trading and putting the long-standing internet infrastructure company directly into one of the biggest capital races in technology.
The deal is remarkable for its size.
It is also remarkable for what it says about the changing AI infrastructure market.
Anthropic, the company behind Claude, has agreed to use Akamai Cloud's distributed infrastructure and software to support its rapidly growing CPU workloads. The agreement can potentially expand by another $9 billion, bringing the total potential commitment to roughly $20 billion.
That puts Akamai in a very different position than the one investors traditionally associated with the company.
Akamai built its reputation around content delivery, cybersecurity and distributed computing. Its infrastructure places computing resources closer to internet users, helping websites and applications load faster and operate more reliably.
Artificial intelligence creates a new use case for that architecture.
As AI applications spread, not every computing task needs to run on the biggest graphics processors used for frontier-model training.
A huge amount of work can also be handled by CPUs — conventional central processing units used for data preparation, orchestration, inference support and many other workloads.
Anthropic's commitment suggests that demand for that capacity is expanding rapidly.
The deal also demonstrates how AI infrastructure is becoming increasingly diverse.
The market often focuses on Nvidia GPUs and giant hyperscale data centers. But a modern AI system requires far more than the accelerators themselves.
It needs networking.
It needs storage.
It needs CPUs.
It needs security.
It needs distributed infrastructure.
And it needs the ability to move data between different parts of the computing stack quickly and reliably.
Akamai is effectively betting that its decades-old edge architecture can become part of that new AI economy.
The financial terms are equally eye-catching.
Akamai said the $11.6 billion commitment will require approximately $5.5 billion in capital expenditures associated with the deal. The company expects to increase its 2026 capital expenditures by about $1.7 billion to secure and pre-purchase critical supply-chain components, including memory.
That means the company is not simply collecting a giant check.
It has to spend heavily to build the infrastructure needed to fulfill the contract.
And Akamai does not expect the agreement to change its 2026 revenue guidance, highlighting that the economic benefit is expected to arrive over several years rather than immediately.
That timing is important for investors.
A contract can look enormous on paper while requiring considerable upfront capital.
The question becomes whether the revenue eventually generated will produce attractive returns after accounting for equipment, financing, energy, operations and other expenses.
Akamai appears confident enough to make the investment.
Investors reacted even more aggressively.
Reuters reported that Akamai shares jumped 22% in extended trading following the announcement.
The reaction makes sense when viewed through the lens of scale.
Before the announcement, Akamai had already disclosed more than $2.8 billion of multi-year Cloud Infrastructure Services commitments across its customer base during 2026.
The Anthropic agreement is several times larger than that entire previously announced figure.
In one transaction, Akamai dramatically increased the visibility of its AI infrastructure business.
There is another unusual feature.
Akamai is giving Anthropic a financial stake in the company.
Under the agreement, Akamai issued Anthropic a warrant for non-voting convertible Series B preferred stock equivalent to about 7.7 million common shares, representing up to approximately 5% of Akamai's common stock outstanding on an as-converted basis.
About 2% is expected to vest in connection with the initial $11.6 billion commitment.
The remaining approximately 3% is linked to an expansion of the relationship of up to another $9 billion.
That structure creates a powerful strategic connection.
Anthropic is not merely a customer.
It becomes potentially one of Akamai's shareholders.
That means both sides have a financial interest in the relationship succeeding.
It is an increasingly familiar pattern in the AI economy.
Infrastructure providers are not just selling compute.
They are increasingly forming deep commercial relationships with the AI companies consuming it.
The model resembles the broader structure of the cloud industry, where major technology companies make enormous infrastructure investments based on long-term commitments from customers.
But the size of today's AI contracts can be extraordinary.
Why?
Because AI companies are consuming computing power at an unprecedented rate.
Anthropic is growing its Claude business while expanding model capabilities and enterprise deployments. More users, more agents and more complex AI applications mean more computation.
That creates a virtuous cycle for infrastructure providers.
More AI adoption generates more workloads.
More workloads require more data-center capacity.
More capacity creates larger long-term contracts.
Those contracts justify additional capital expenditure.
And that new infrastructure enables even more AI adoption.
The cycle can become enormous.
It can also become financially risky.
Akamai's $5.5 billion capital-spending commitment shows that the infrastructure side of the AI boom is not free.
Someone has to buy servers.
Someone has to acquire memory.
Someone has to secure power.
Someone has to finance the buildings and networking equipment.
This is why the Akamai-Anthropic deal matters beyond the two companies.
It is another example of how the AI boom is spreading into the credit, infrastructure and capital-spending markets.
Investors are no longer simply betting on which company will build the best model.
They are also betting on the companies that supply the computing power required to run those models.
Akamai's distributed network could become particularly useful because AI applications are not necessarily confined to massive centralized clusters.
Many applications need low-latency processing closer to users.
Akamai's edge-computing heritage gives it a potentially differentiated position in that market.
The company's global network spans thousands of points of presence, allowing workloads to be placed closer to end users. Akamai says its cloud platform is designed to support AI applications from the core to the edge while providing security and reliability.
Anthropic's decision to commit at this scale therefore sends a signal about the types of infrastructure the AI industry may need.
AI is not simply one giant server farm.
It is an ecosystem.
And Akamai wants to own a much larger slice of it.
Still, investors will need to watch the economics closely.
The $11.6 billion headline is impressive, but the company has significant capital requirements before it can fully deliver on the agreement.
Akamai's own filings warn that the deal depends on supply-chain availability, customer demand, infrastructure performance and the ability to execute the expansion.
The company also faces the normal risks of any capital-intensive business, including competition and fluctuating technology costs.
Yet the market's immediate reaction tells its own story.
For years, Akamai was known primarily as a behind-the-scenes company — important, but rarely at the center of technology headlines.
Anthropic has changed that.
With one $11.6 billion commitment, AI has effectively moved Akamai into the spotlight.
The company now has to prove that its distributed cloud can support one of the world's fastest-growing AI businesses at enormous scale.
If it does, the deal could become a defining example of how traditional internet infrastructure gets transformed by the AI era.
The big AI winners may not all be the companies building the models.
Some could be the companies building the enormous machines underneath them.
Akamai has just received a very large vote of confidence from Anthropic.
Now comes the expensive part: building the infrastructure to earn it.