MW Anthropic's meteoric rise shocked the market - but the AI crown remains up for grabs
By Christine Ji
Anthropic has emerged from obscurity as the industry's most formidable AI challenger. But maintaining that momentum will be a tall order, as OpenAI fights to hold its lead.
Anthropic, co-founded by Dario Amodei $(R)$, recently surpassed $19 billion in run-rate revenue, closing in on Sam Altman's (L) OpenAI, with run-rate revenue of $25 billion.
A year ago, Anthropic was a niche artificial-intelligence lab in the shadow of OpenAI. Now, Anthropic's Claude sits at the top of the App store in 16 different countries.
The balance of power among AI frontier labs is undergoing a seismic shift yet again, as OpenAI finds itself fending off challengers in a race that it once dominated alone. Anthropic, fueled by momentum from Claude Code and the recent publicity from its standoff with the Pentagon, has aggressively grown revenue and seized mindshare.
But those rushing to declare Anthropic the new king of AI shouldn't celebrate prematurely.
In fact, "I don't actually know why that would be the assumption," Wharton professor Ethan Mollick, who studies the impact of AI on work and education, told MarketWatch. Despite recent headlines, the AI industry's heavyweights remain locked in a high-velocity arms race with no end in sight.
The competition between OpenAI, Anthropic and Alphabet's $(GOOGL)$ $(GOOG)$ Google Gemini remains "neck and neck," according to Mollick. On Thursday, OpenAI launched its newest GPT-5.4 model, featuring more advanced reasoning, coding and document-editing abilities.
GPT-5.4 is "probably the best model in the world right now," Mollick said. But "all three of the major model makers seem to be in the race of releasing models that keep one-upping each other. We are seeing steady and exponential increases in AI ability over time," he added.
Still, the rise of Anthropic would have been practically unthinkable just a few months ago. In a report from last September, JPMorgan analyst Brenda Duverce pointed out that Claude made up less than 1% of mobile downloads and had less than 2% share of daily AI web traffic. Opting to pursue enterprise customers, Anthropic seemed woefully behind OpenAI in the consumer realm - until now.
More: How can Anthropic stand out in the AI wars? I went to a Greenwich Village pop-up to find out.
Last Friday, OpenAI's decision to strike a deal with the Pentagon triggered a massive wave of public backlash and resulted in ChatGPT "uninstalls" spiking 295% over the weekend, according to the market-intelligence firm Sensor Tower. As of Thursday, Claude recorded over 1 million daily signups.
Don't miss: Anthropic has 'no choice' but to fight Pentagon's 'supply-chain risk' designation in court, CEO says.
Anthropic recently eclipsed $19 billion in run-rate revenue, Bloomberg reported this week, up from $14 billion less than a month ago and $9 billion at the end of 2025. That puts Anthropic within striking distance of OpenAI, which has a $25 billion run-rate revenue, according to a Wednesday report from The Information.
Read: Facing backlash, OpenAI's Sam Altman says he made a 'sloppy' mistake in Pentagon deal
Such upsets are becoming a hallmark of the AI landscape as frontier labs race to achieve artificial general intelligence - technology with the rumored ability to do anything a human can and unlock trillions in economic value. At the bleeding edge of AI, any new feature has the potential to evaporate an incumbent's lead overnight.
It's a dynamic that Google is intimately familiar with. Last August, Gemini's viral Nano Banana image-editing feature catapulted the chatbot to the top of the App Store. Investors, once convinced that ChatGPT had sent Google search to death's door, began pouring money into the tech conglomerate.
Mounting concerns about OpenAI's ability to monetize its $1 trillion of circular deals rattled investors and further solidified Gemini's lead as investors lauded the tech conglomerate's full-stack AI infrastructure and unparalleled distribution advantage. The momentum delivered Alphabet the title of best-performing "Magnificent Seven" stock in 2025 and led to OpenAI declaring an emergency "code red" last December.
Anthropic poses a different threat to OpenAI than Google. The startup was forged during a 2021 executive exodus at OpenAI, when Dario Amodei, the current Anthropic CEO, and a group of researchers split off to build a rival focused on AI safety.
In the years since, Anthropic has powered its rise with significantly fewer financial resources than OpenAI. OpenAI's latest $110 billion funding round values the ChatGPT creator at a whopping $730 billion pre-money, compared with Anthropic's $30 billion fundraise and $380 billion post-money valuation last month. Both companies are reportedly eyeing public market debuts in 2026.
Read: As ChatGPT turns 3, here's what's crashing the party
While recent controversies with the Pentagon have boosted Anthropic's name recognition and tarnished OpenAI's reputation, Stefan Slowinski, global head of software research at BNP Paribas, doesn't think the tailwind is "sustainable" for Anthropic. As of Thursday, Anthropic is reported to be back in talks with the U.S. government.
"I would expect Anthropic to work with the U.S. government in some form," Slowinski said. "We see all major technology vendors doing that," he added, highlighting that Anthropic's primary cloud partner Amazon.com (AMZN) works closely with various government entities. Ultimately, the political controversy won't be a significant factor in enterprises' long-term IT spending decisions, according conversations Wharton's Mollick had with corporate leaders.
See more: Trade Desk's stock soars, as a potential OpenAI partnership gave investors reason to cheer
Among investors, there is some concern that OpenAI is spreading itself too thin, Slowinski said, as it pursues consumer and enterprise end markets as well as advertising, e-commerce and wearables. The fear is that OpenAI has entered into a multi-front war, risking a scenario where it loses its consumer edge to Google while ceding the high-margin enterprise crown to a more focused Anthropic.
Anthropic has indeed proven to be a formidable player in the enterprise space as its recent product launches have spurred a brutal selloff among software stocks. From Claude-powered Excel plugins to its ability to rewrite decades-old COBOL infrastructure, Anthropic's product suite looks capable of rendering a significant portion of the enterprise software market obsolete.
Read on: IBM's stock heads for worst month in 34 years - and Anthropic is partly to blame
But Claude Code won't be sufficient for Anthropic to maintain its edge going forward.
Both Mollick and Slowinski believe that Anthropic's early-mover advantage in agentic coding is already being eroded by OpenAI's own coding agent, Codex, which has seen significant uptake since its launch in April of last year. According to The Information, weekly average users of Codex have quadrupled since the start of the year.
"It seems like [OpenAI] may have caught up in terms of capabilities on the coding side with Anthropic," Slowinski told MarketWatch. Slowinski also noted that financial analysis appears to be a core competency for OpenAI's newest GPT-5.4 model, marking another challenge to Anthropic's enterprise strongholds.
"From a capability standpoint, they probably are perceived to be in a better place than they were at the end of last year when they had that 'code red,'" Slowinski said of OpenAI.
Investors shouldn't even take the current dominance of OpenAI, Anthropic and Google as an absolute, as Slowinski cautioned that a dark-horse challenger could still emerge to disrupt the hierarchy in the coming months.
He pointed out that Meta Platforms (META) has been building its Superintelligence Lab with two models - dubbed "Mango" and "Avocado" - rumored to come out later this year. Elon Musk has merged xAI with SpaceX, strengthening the financial position of xAI. And Slowinski pointed to China's DeepSeek, which is also rumored to release a new model this year.
It's a stark reminder that in the current climate, an AI lab's technical advantages are increasingly short-lived.
-Christine Ji
This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
March 06, 2026 07:01 ET (12:01 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments