Some early investors in OpenAI are raising concerns about the company's $852 billion valuation, citing a series of strategic pivots that they say make the startup appear unfocused, according to a recent report. The criticism comes as OpenAI shifts its focus toward higher-margin enterprise sales, an area where it lags behind rival Anthropic, leaving it vulnerable to competition from both Anthropic and Google.
“You have ChatGPT, a 1 billion-user business growing 50-100 per cent a year, what are you doing talking about enterprise and code? It’s a deeply unfocused company,” an unnamed early backer told the Financial Times. This sentiment reflects a broader dissatisfaction among some investors who feel that OpenAI is squandering its dominant position in the consumer AI space by chasing multiple directions at once.
Strategic Pivot Raises Eyebrows
OpenAI’s recent moves have been anything but predictable. The company shut down its video generation tool Sora, a move that eliminated a $1 billion investment from Disney and dashed hopes for a creative content pipeline. It also scrapped plans for an “adult” chatbot, drastically pared back an investment deal with Nvidia, and halted development of a $30 billion data center in the UK while pausing expansion of a site in Abilene, Texas. These decisions have left analysts and investors questioning the company’s long-term roadmap.
Instead, OpenAI has begun pushing its Codex coding tool to businesses, putting it in direct competition with Anthropic’s Claude. Codex, initially developed as a research project, is being repositioned as a developer platform for enterprises looking to automate code generation and review. However, Anthropic has already secured partnerships with several large corporations, giving it an early lead in the enterprise coding space.
Investor Sentiment and Valuation Concerns
An investor who has backed both OpenAI and Anthropic noted that investing in OpenAI’s latest funding round would require assuming an IPO valuation of $1.2 trillion or more—a figure that is increasingly hard to justify given the cheaper proposition of buying into Anthropic, valued at $380 billion. This disparity highlights a growing disconnect between OpenAI’s astronomical valuation and its tangible business outcomes.
Another critic pointed to OpenAI’s purchase of tech talk show TBPN as “a distraction,” arguing that the acquisition does little to further the company’s core mission of advancing safe artificial general intelligence. “It feels like they are trying to be everything to everyone,” said a third investor who wished to remain anonymous. “When you have a product like ChatGPT that is printing money, you double down on that, not diverge into media and coding tools.”
Competitive Landscape: Anthropic and Google
The market for AI-powered enterprise solutions is heating up, with Anthropic emerging as a formidable challenger. Anthropic’s focus on safety and reliability has won over clients in regulated industries such as healthcare and finance, while its Claude model is praised for its nuanced handling of complex tasks. Google, too, is aggressively pushing its Gemini platform into enterprise settings, leveraging its vast cloud infrastructure and existing relationships with businesses.
OpenAI’s strength, however, lies in its computing resources. The company has secured massive clusters of Nvidia GPUs and has access to Microsoft’s Azure cloud infrastructure, giving it a significant edge over Anthropic in terms of raw processing power. CFO Sarah Friar defended the company’s direction, stating that the large size of its recent funding round—which raised over $10 billion—demonstrates confidence from major investors. “We are building for the long term,” she said. “Enterprise and coding are natural extensions of our core product, and we have the resources to execute.”
The Nvidia Partnership and Data Center Deals
OpenAI’s relationship with Nvidia has been a cornerstone of its compute strategy. The company initially planned to invest billions in specialized Nvidia hardware for data centers, but that deal was drastically scaled back in recent months. Sources suggest that OpenAI is now focusing on optimizing its existing infrastructure rather than expanding aggressively. This shift has led to the cancellation of the UK data center project, which was intended to support European customers and comply with local data regulations.
The Abilene, Texas site has also stalled, with plans to extend capacity put on hold. While the company says it is reevaluating its global footprint, critics argue that these retreats signal uncertainty about demand. “Building data centers is a signal of long-term commitment,” said Jai Das, president of Sapphire Ventures, who is not an investor in either company. “When you pull back, it raises questions about your conviction.”
Codex and the Enterprise Push
Codex is now being marketed as a turnkey solution for software development teams. The tool can generate entire codebases from natural language descriptions, debug existing code, and even suggest optimal architectures. OpenAI is offering tiered pricing for Codex, with enterprise plans starting at $100,000 per year. This puts it in direct competition with Anthropic’s Claude for Developers, as well as GitHub Copilot, which uses a different AI model.
However, early feedback from enterprise clients has been mixed. Some praise Codex for its ability to handle large projects but complain about its high cost and occasional inaccuracies. Others note that Anthropic’s Claude has a reputation for producing more secure code, a critical factor for financial and healthcare applications. The battle for the enterprise will likely be decided by which company can deliver the most reliable and secure solution at the right price.
Infrastructure Lead and Market Confidence
Despite the criticisms, OpenAI maintains a significant lead in procuring computing resources. The company has partnerships with not only Nvidia but also startup firms specializing in AI chips. Its access to Microsoft’s Azure gives it a cloud platform that is deeply integrated with the Microsoft 365 suite, offering potential synergies that competitors cannot match. “OpenAI has the raw horsepower,” said one industry analyst. “But horsepower alone doesn’t win races if the driver keeps changing lanes.”
Friar’s comments about investor confidence were echoed by a few deep-pocketed backers who see the long-term potential. One of them argued that the current criticisms are short-sighted: “Companies like OpenAI are building the infrastructure for the next decade. You can’t judge them quarter by quarter.” Nevertheless, the noise from early investors suggests that patience is wearing thin.
The Netscape of AI?
Jai Das’s comparison of OpenAI to Netscape—the browser giant that dominated the late 1990s before being overtaken by Microsoft and eventually sold to AOL—has resonated with some observers. Netscape’s failure to monetize its early lead and its subsequent flailing into unrelated businesses (like internet portals) is seen as a cautionary tale. “OpenAI has the eye of the tiger, but it needs to keep its focus,” Das said.
Others, however, argue that the analogy is flawed because AI is a foundational technology with multiple killer applications. “Netscape was just a browser. OpenAI is creating an entirely new way for humans to interact with machines. That distinction matters,” said an industry veteran. For now, the company remains the leader in generative AI, but its strategic zigzags are giving competitors an opening.
Source: Silicon UK News