“Code red”: OpenAI rethinks its all-things-AI strategy
Internal warning signals push the company toward enterprise customers and away from consumer sprawl.
OpenAI’s Pivot: The AI company is planning to reorganize its business strategy, shifting its focus toward coding tools and enterprise customers at the expense of the consumer market that has driven its growth to date. The move comes in response to Anthropic’s success in the enterprise segment and OpenAI’s ongoing struggle to establish a sustainable business model that justifies its soaring costs.
“A strategy of ‘doing everything all at once’ puts us on the defensive,” senior company officials said, according to a report in The Wall Street Journal.
OpenAI’s products, led by ChatGPT, have achieved massive adoption. By February, the chatbot had reached 900 million weekly users, and in January the company reported that its 2025 revenue had grown 233% to $20 billion. However, expenses continue to far exceed revenues. The company is estimated to be losing tens, if not hundreds, of billions of dollars annually, largely due to the massive infrastructure required to train and operate advanced AI models.
In an effort to close this gap, OpenAI has experimented with new business models beyond monthly subscriptions, which only a minority of users pay for, including the controversial idea of integrating advertisements into ChatGPT. Increasingly, however, the company appears to be looking toward Anthropic’s approach for guidance.
Anthropic has focused on enterprise customers, who are willing to pay significantly more for technologies that deliver measurable efficiency gains. Its products, such as Claude Code, have gained traction, alongside additional capabilities for its Claude chatbot in areas such as data analytics, legal workflows, and cybersecurity. The rollout of these tools has coincided with declines in the shares of traditional software providers offering similar services.
OpenAI is now attempting a similar shift. According to the Journal, the company’s chief application officer, Fidji Simo, told employees last week that CEO Sam Altman and research chief Mark Chen are reassessing which areas to prioritize, and which to scale back.
“We can’t let this moment slip away because of distractions from fringe activities,” Simo said. “We have to double down on productivity in general, and on enterprise productivity in particular.”
Throughout 2025, OpenAI expanded aggressively across multiple fronts, launching products ranging from the video-generation model Sora to the AI browser Atlas, as well as integrating e-commerce capabilities into ChatGPT. The company also acquired a startup founded by Jony Ive, aiming to develop AI-based hardware.
Altman previously described this approach as “a bet on a series of startups,” helping reinforce OpenAI’s reputation as a leading innovator. But as costs rise and concerns grow about an overheated AI market, pressure is mounting to deliver a clearer and more sustainable business model, particularly as the company considers a potential IPO later this year.
By contrast, Anthropic’s more focused strategy, eschewing areas like image and video generation, has positioned it as a leader in monetization. According to the Journal, Simo told employees that Anthropic’s progress should serve as a wake-up call, urging OpenAI to reassert its leadership among developers and enterprise customers.
Current and former employees told the WSJ that OpenAI’s “do-it-all” strategy created internal confusion and inefficiencies. Computing resources were frequently reallocated between teams at the last minute, and the company’s organizational structure became increasingly complex. For example, the Sora team was placed within the Research division, despite being responsible for launching one of the company’s most prominent products.
In September, OpenAI launched a standalone Sora app that combined video generation with TikTok-style social features. Although the app initially reached the top of Apple’s App Store rankings, its momentum faded in subsequent months, and the company is now considering integrating its video capabilities directly into ChatGPT.
Meanwhile, Anthropic has strengthened its position with updates to Claude Code, which has become a preferred tool for many software engineers. OpenAI is attempting to close the gap with a revamped version of its coding tool, Codex, and with GPT-5.4, which is optimized for enterprise use.
The company has also benefited from Anthropic’s reluctance to work with the Pentagon, opening opportunities for OpenAI to pursue defense-related contracts. Still, executives increasingly acknowledge the need for a sharper strategic focus.
“We are very much acting as if it’s a code red,” Simo told staff according to the Journal. “I don’t think necessarily declaring codes for everything makes a ton of sense.”
