Anthropic CEO Dario Amodei said Wednesday that the artificial intelligence company underestimated the scale of demand for its Claude models so dramatically that even aggressive growth projections failed to prepare it for the surge now straining its infrastructure.
Speaking at Anthropic’s developer conference in San Francisco, Amodei revealed that the company planned for tenfold growth, only to see revenue and usage explode by roughly 80 times in the first quarter on an annualized basis.
“That is the reason we have had difficulties with compute,” Amodei said. “We are working as quickly as possible to provide more.”
The remarks offered one of the clearest illustrations yet of the extraordinary pressure sweeping through the AI industry, where surging adoption of generative AI tools is rapidly outpacing the world’s ability to build enough data centers, power systems, and advanced chips to support them.
Anthropic’s sudden expansion also highlights how the AI race is evolving from a battle over algorithms into a full-scale infrastructure war centered on compute capacity, electricity access, and semiconductor supply chains. The company’s growth has been fueled largely by explosive demand for its Claude family of AI models, particularly Claude Code, which has emerged as one of the fastest-growing AI programming tools in the software industry since its launch last year.
“Software engineers are the ones who are fastest to adopt new technology,” Amodei said on stage. “It’s a foreshadowing of how things are going to work across the economy, and how the economy is going to be transformed by AI.”
The scale of the growth is now forcing Anthropic into increasingly aggressive efforts to secure computing power. Hours before Amodei’s appearance, the company announced a major agreement with Elon Musk’s SpaceX to utilize the full compute capacity of the Colossus 1 data center in Memphis, Tennessee.
Under the deal, Anthropic will gain access to more than 300 megawatts of computing capacity, a massive allocation that underscores how AI firms are now competing for power infrastructure on a scale previously associated with heavy industry.
The agreement is particularly striking because SpaceX owns rival AI lab xAI, now known as SpaceXAI, making the partnership an unusual alliance between competitors in one of the world’s most aggressive technology races.
The deal also marks a shift underway across Silicon Valley, where access to compute has become as strategically important as access to capital. Companies developing frontier AI systems increasingly face constraints not from talent or software innovation, but from shortages of GPUs, electricity, cooling systems, and data center construction capacity.
Industry analysts have warned that the global AI boom is beginning to reshape energy markets, real estate development, and industrial supply chains as hyperscalers and AI startups rush to secure long-term infrastructure.
Anthropic itself acknowledged last month that demand for Claude had created “inevitable strain on our infrastructure,” affecting reliability and performance, especially during peak usage periods.
The company has responded with a string of major compute agreements, including a multibillion-dollar partnership with Amazon, which has become one of Anthropic’s most important strategic backers.
The rapid expansion has also transformed Anthropic into one of the world’s most valuable private technology companies.
The startup is reportedly in discussions with investors to raise fresh funding at a valuation of approximately $900 billion, according to CNBC. Such a valuation would place Anthropic above OpenAI, marking a stunning rise for a company founded only a few years ago by former OpenAI researchers.
The investor appetite reflects growing belief across financial markets that generative AI could fundamentally reshape global industries ranging from software engineering and finance to healthcare, manufacturing, and defense.
At the same time, Anthropic’s rise has intensified political and regulatory scrutiny. The company remains locked in a contentious dispute with the U.S. government and the Pentagon after the Defense Department labeled Anthropic a “supply chain risk” in March and barred the use of its products across military networks and contractors.
The blacklisting triggered legal and political fallout, particularly because Anthropic’s tools had already become deeply embedded across parts of the defense ecosystem. Pentagon officials and contractors have reportedly resisted efforts to phase out Anthropic products, viewing the company’s AI systems as technologically superior to many rivals.
The conflict also highlights growing tensions between Washington’s national security concerns and the private sector’s breakneck AI expansion. Anthropic’s advanced cyber-focused AI systems, especially its controversial Mythos model, have triggered fears among U.S. officials that such tools could dramatically increase offensive hacking capabilities if deployed irresponsibly.
Even as those concerns deepen, demand for Anthropic’s products continues accelerating at a pace that Amodei himself suggested may be unsustainable.
“The current level of growth is just crazy,” he told attendees. “Too hard to handle.”
He added that he hopes the company eventually sees “more normal” expansion, though current market dynamics suggest the global AI boom is still in its early stages.
Anthropic’s infrastructure challenges mirror wider strains emerging across the AI industry. Microsoft, Amazon, Google, Meta, and OpenAI are collectively spending hundreds of billions of dollars on data centers, networking equipment, and AI chips as they race to meet surging enterprise and consumer demand.
That spending frenzy has become a major driver of global semiconductor markets, electricity demand, and construction activity, while also raising concerns among economists that AI infrastructure investment itself could contribute to inflationary pressures.
For now, however, investors continue rewarding companies tied to the AI ecosystem, betting that the technology’s transformative potential outweighs concerns about soaring valuations, infrastructure bottlenecks, and regulatory risks.
Anthropic’s explosive growth has now become one of the clearest signals yet that the global AI arms race is accelerating faster than even many of its architects expected.
