Elon Musk said the words in public, without prompting and without qualification: he misjudged Anthropic. The remark, made on July 9, was a rare admission from an executive who rarely concedes ground, and it landed in the same week as a report that put the company’s revenue on a trajectory that no one in the industry predicted a year ago.
The numbers come from NextBigFuture, which reported that Anthropic’s annualized revenue run rate has reached roughly $69 billion, up from under $10 billion at the end of 2025. If the figure is accurate, it represents one of the fastest revenue ramps in the history of the software industry, and it would put Anthropic, a company founded in 2021 as a safety-focused research lab, among the largest sellers of AI in the world.
The revenue, if the run rate holds, comes from a business built on a different model than its rival’s. Where OpenAI’s growth was propelled by a consumer chatbot, Anthropic’s acceleration has been driven by enterprise contracts, API usage and the deployment of its models inside regulated industries, where the company’s emphasis on reliability has been a selling point rather than a slogan. The mix matters for the quality of the revenue: recurring contracts and usage commitments are worth more to investors than consumer seats, and the composition of the $69 billion is as important as its size.
The valuation math is the part investors will scrutinize. Anthropic was last reported to be worth about $965 billion, and against a $69 billion revenue run rate that implies a price-to-sales ratio in the neighborhood of 14 times, a multiple that, while rich, is not extreme by the standards of the AI market. The arithmetic suggests that the market has already begun pricing Anthropic as a mature commercial enterprise rather than a research bet.
His relationship with Anthropic has been tangled from the start. Musk helped found OpenAI in 2015, and he was an early financial backer of Anthropic before the two organizations diverged, and his public posture toward the company turned hostile as it grew. He has called its approach to safety overly cautious and its leadership ideological, and his remarks on July 9, acknowledging that he misjudged the company, amount to a rare public reassessment of a competitor he had spent years dismissing. The industry read the moment as a marker of how far the perception of Anthropic has traveled.
Musk’s history with the company makes the admission pointed. He was a founding figure of OpenAI, Anthropic’s chief rival, and he has spent years describing Anthropic as an over-regulated, safety-obsessed competitor that would be outflanked by faster, more aggressive labs. His public reversal, acknowledging Anthropic as one of the leading forces in AI, was read across the industry as a recognition that the safety-first positioning, far from being a handicap, became a commercial advantage as enterprises sought AI vendors they could trust.
The growth story behind the numbers is a study in enterprise strategy. Where OpenAI built its franchise on a consumer product, Anthropic concentrated on selling its models to businesses, focusing on the industries, banking, health care, legal and government, where reliability and guardrails are procurement requirements rather than marketing talking points. The approach produced slower headlines and steadier contracts, and the revenue run rate, if accurate, shows the compounding effect of that choice.
The shift in perception is generational within the industry. Two years ago, the standard narrative divided AI companies into safety-first labs and commercial-first companies, with the implicit judgment that safety orientation was a constraint on growth. Anthropic’s numbers suggest the opposite ordering: the lab that built its brand on restraint is now outgrowing rivals that built theirs on speed, and the industry’s internal map of who is a safe bet and who is a risk has been redrawn accordingly.
The IPO question now dominates conversations about the company. Prediction markets, which have been reasonably accurate on AI-sector events, put the probability of an Anthropic listing within 2026 at above 70 percent, and bankers said the revenue trajectory, if it holds, would support one of the largest technology listings in years. The company has said nothing official, and people familiar with the matter said the timing remains an internal decision, but the market’s expectation is set.
The stakes extend beyond Anthropic itself. A public Anthropic at a $69 billion run rate would reset the valuation benchmarks for every private AI company, including OpenAI, and it would give investors a liquid way to express views on the sector’s economics. The listing would also test the market’s appetite for AI companies at scale, after a period when the sector’s valuations have swung with every quarter of model news.
Musk’s acknowledgment, in that context, was the industry catching up with its own data. The company he dismissed as over-regulated is now the one the market is pricing for a record listing, and the reversal carries a lesson that executives across the sector are repeating in internal meetings: in AI, the fastest path is not always the one that wins. Anthropic’s run rate, if it holds, is the evidence, and the market is waiting for the financial statements that would confirm it.
