Anthropic and Google pressure: Why OpenAI is urgently changing its strategy

OpenAI, valued at $852 billion, is undergoing a period of strategic change. As reported by Reuters, following a record-breaking $122 billion funding round, the company is making a sharp pivot, abandoning some ambitious projects and focusing on developer tools and business solutions to justify its massive valuation and stay competitive.

Expansion and Loss of Focus

Over the past six months, OpenAI has revised its product roadmap twice. The first catalyst was the launch of Google's new Gemini model, which prompted the company to declare a "red alert." The next blow came with the emergence of Anthropic's Claude Code—a developer tool that quickly gained popularity and forced OpenAI to urgently reallocate resources in favor of its own product, Codex.

These events revealed that the company's previously sprawling strategy—featuring numerous disparate projects—was creating internal conflicts over talent, computing power, and other resources. OpenAI realized that expanding into too many directions simultaneously increased the risk of losing focus, despite its ambitious goal of creating artificial general intelligence (AGI).

Closing Sora and Partner Reactions

One of the first and most visible results of this pivot was the closure of the AI video application Sora. The news came as a shock to many employees and partners, including Disney, which had signed a $1 billion agreement with OpenAI for the Sora project in December of last year. According to Reuters, Disney and OpenAI employees worked late into the night on Sora the day before the application's closure, unaware of the company's decision.

This episode highlights how abrupt and unpredictable OpenAI's pivots can be. On one hand, it is a typical startup story of redirecting resources toward more promising areas. On the other hand, it demonstrates the risks for partners and employees when a company changes priorities without long-term preparation.

Betting on Codex and Business Solutions

Today, OpenAI is betting on developer tools and enterprise solutions. At the center of the new strategy is Codex and the development of a super-app that will integrate ChatGPT with other company products. "Users don't want disparate tools; they need a unified system that understands intent, executes actions, and works with applications, data, and workflows," the company states.

Fiody Simo, one of OpenAI's leaders, emphasizes: "When new directions start working, as we see with Codex, it's important to double down on them and avoid distractions. We are glad to be capitalizing on this moment."

Ambitions vs. Reality

Six months ago, OpenAI was a completely different company. It was developing multiple directions simultaneously: AI video, consumer devices, e-commerce, and advertising for ChatGPT. The company entered into major partnerships, including integrating ChatGPT into Mattel's Barbie, discussed infrastructure deals worth over $1 trillion, and prepared for a trillion-dollar IPO.

Sam Altman claimed that these disparate projects were part of an overall strategy to create AGI. However, pressure from competitors, growing investor demands, and real infrastructure limitations forced OpenAI to adjust its ambitions.

Infrastructure Constraints and Competition

The AI industry faces not only competition but also physical world constraints. According to Morgan Stanley, the largest technology companies—Amazon, Microsoft, Alphabet, and Meta—will spend approximately $630 billion on AI infrastructure in 2026, representing about 2.2% of the U.S. GDP. Most of these investments go toward servers and GPUs, but buildings, networking equipment, security, and electricity are also required.

Access to stable energy and infrastructure has become a bottleneck. Connecting to the public power grid in major cities can take up to ten years, while remote regions often suffer from a shortage of skilled labor. Self-generating energy faces equipment shortages, such as gas turbines, which will not be available until after 2029.

Conclusion

Amidst accelerated competition and massive infrastructure costs, OpenAI must find a balance between ambitious projects and the need to focus on priority areas. The company's pivot toward Codex and business solutions shows that even the world's largest and most expensive AI company cannot afford to scatter its resources without risking the trust of partners and investors.