Anthropic is heading toward its first profitable quarter, a sharp turn for a company that only recently looked trapped between explosive demand and punishing compute costs.
The clearest version of the story is simple: Anthropic is telling investors it expects to turn an operating profit in the second quarter, even as it keeps spending heavily to feed Claude's growth. The Wall Street Journal first reported that the company expects about $559 million in operating profit on roughly $10.9 billion in revenue, and Reuters later reported the same figures in a fresh look at Anthropic's compute commitments.
That matters because Anthropic has spent the past two years being treated as the AI startup most likely to prove that safety and scale can coexist. The company built its reputation on a more cautious approach than some rivals, then backed that posture with a product that enterprise buyers clearly wanted. Now the numbers suggest that discipline may be translating into something even more valuable than brand trust: actual operating leverage.
The most important driver here is not consumer hype, it is enterprise adoption. Reuters reported that Bristol Myers Squibb is making Claude available to more than 30,000 employees as part of a push to speed up drug discovery, development, and delivery. TechCrunch, meanwhile, reported that Anthropic has overtaken OpenAI among verified business customers on Ramp's spending data, which is the kind of signal that usually tells you where the serious revenue is coming from.
That is the broader change investors are watching. Anthropic has not simply sold a clever chatbot. It has been converting Claude into an operating system for white-collar work, one contract at a time, through APIs, enterprise subscriptions, and workflow automation. In a market where a lot of AI usage still looks experimental, that is the difference between a demonstration and a business.
The Bristol Myers deal is especially useful as a case study because it shows how AI vendors are moving beyond general productivity claims. If Claude is useful in drug discovery, internal knowledge work, and regulated environments, then the company has a path to revenue that is both sticky and high-value. That is how a startup moves from growth story to financial story.
Profit now, pressure later
The timing, though, is awkward in a way that only the AI market can produce. Reuters reported that Anthropic will pay SpaceX $1.25 billion per month for compute through May 2029, with the details emerging from SpaceX's S-1 filing. The agreement reportedly gives Anthropic access to 300 megawatts of compute from the Colossus 1 data center near Memphis, and it may be terminated by either side with 90 days' notice.
That is a huge bill, and it cuts straight to the core of the AI business model. The most ambitious model makers are no longer limited by demand, they are limited by the cost of running the models at scale. Anthropic's apparent move into profit does not erase that problem. It just shows that for now, revenue is rising faster than the infrastructure tab.
The catch is that this may not last. Reuters noted that Anthropic may not stay profitable for the full year because of its coming compute and training expenses, which is exactly why the Q2 milestone should be read carefully. This is not a clean break into durable profitability. It is a snapshot of a company growing quickly enough to absorb massive costs, at least for the moment.
Still, that snapshot matters. A profitable quarter changes the conversation around valuation, fundraising, and eventually public markets. It gives Anthropic a new argument in a sector where many firms still rely on the promise of future monetization. If the company can show that safety-focused AI can also produce real operating profit, it strengthens the case that enterprise buyers will pay for reliability, not just novelty.
It also changes how Anthropic is compared with OpenAI. Reuters reported that the company's revenue gap with OpenAI has narrowed, and the latest profitability story suggests the market is no longer asking which lab has the loudest product launches. It is asking which one can turn demand into a repeatable business without burning through every dollar it makes.
That is why this moment feels bigger than a single quarter. Anthropic is not just chasing growth anymore. It is testing whether frontier AI can become a normal company, with margins, constraints, and the possibility of an IPO or large secondary sale hanging just over the horizon.
Also read: White House AI cyber order could tighten rules for federal deployments • SpaceX's IPO filing exposes xAI's huge cash burn • Jensen Huang is widening Nvidia's target far beyond chips
