OpenAI's Enterprise Share Fell From 50% to 27% in Two Years
OpenAI, Once the Pace-Setter, Is Now Chasing Anthropic By Sky | type0 OpenAI built the template.

image from Gemini Imagen 4
OpenAI built the template. Now it is living in someone else's frame.
The company that launched the AI boom with ChatGPT is executing a strategic pivot that its own leadership once reserved for rivals: a deliberate, urgent chase. According to reporting by the Wall Street Journal and confirmed by sources across the AI industry, OpenAI's top executives are finalizing plans to refocus the company around coding tools and enterprise clients, cutting what one insider described as side quests in favor of a narrower, sharper bet.
The proximate cause is Anthropic.
Fidji Simo, OpenAI's chief of applications, told employees in an all-hands meeting that Anthropic's rise should serve as a wake-up call, according to remarks reviewed by the Journal. "We cannot miss this moment because we are distracted by side quests," Simo said. "We really have to nail productivity in general and particularly productivity on the business front." CEO Sam Altman and chief research officer Mark Chen are actively reviewing which areas to deprioritize, with staff expected to be notified in the coming weeks.
The numbers explain the urgency.
Anthropic holds approximately 40% of the enterprise AI market, according to a December report from VC firm Menlo Ventures. OpenAI's share has collapsed from 50% in 2023 to 27% at the end of 2025. On the coding side specifically, the gap is stark: Anthropic's Claude Code accounts for nearly a fifth of its total business — more than $2.5 billion in annualized revenue, the company said in February. OpenAI's Codex was generating just over $1 billion in annualized revenue by the end of January, according to a person with direct knowledge of the matter, speaking to Wired.
One product tells the whole story.
Sora, OpenAI's video generation tool, briefly hit No. 1 on Apple's App Store after its September launch. Usage flatlined within weeks. OpenAI is now folding Sora's capabilities back into ChatGPT rather than maintaining it as a standalone product — an implicit admission that the separate app was a dead end. The Atlas web browser, another sideline project, is being deprioritized. Hardware and e-commerce features for ChatGPT are under review.
Altman had previously described the scatter-shot approach as "betting on a series of startups" inside the company. That bet appears to have cost OpenAI its enterprise lead at precisely the moment when enterprise customers were becoming the industry's most reliable revenue source.
The pivot has a product shape. OpenAI is planning a desktop superapp that would unify ChatGPT, its coding platform Codex, and a browser into a single application — a consolidated interface for the tools Anthropic has been winning with individually. Fidji Simo has publicly highlighted Codex's recent momentum: the product now has over two million weekly active users, roughly four times the figure at the start of the year.
The competitive dynamic has begun to shape capital structures. Reuters reported this week that OpenAI is in advanced talks with private equity firms including TPG, Bain Capital, Advent International, and Brookfield Asset Management to form a joint venture that would distribute its enterprise tools across their portfolio companies. The proposed deal carries a pre-money valuation of approximately $10 billion, with the PE firms committing roughly $4 billion. TPG would anchor the deal. Notably, OpenAI is offering preferred equity — a senior class of ownership that prioritizes investor returns and limits downside. Anthropic, by contrast, is offering common equity in its own PE venture, a less protected position that reflects different risk appetites.
Both companies are racing toward public offerings. The private equity partnerships represent, in part, a bid to demonstrate enterprise traction that public markets will want to see.
The Wired profile of OpenAI's internal dynamics, based on interviews with more than 30 current employees and leaders, paints a picture of a company that had the coding insight early and did not pursue it. OpenAI demonstrated a code generation tool called Codex in 2021 — two years before ChatGPT existed — and executives at the time recognized it as a potential cornerstone. The team was then reassigned to work on DALL-E 2 and GPT-4. When ChatGPT arrived and exploded in popularity, the consumer moment seemed to validate the broader product play.
Anthropic took a different path, staying focused on coding capabilities even as the industry chased multimodal splash.
Whether OpenAI can close the gap before both companies go public remains an open question. What is clear is that the company that defined the AI era is no longer defining its next chapter.

