Sam Altman halted other projects to save ChatGPT

OpenAI is suspending work on side projects and focusing resources on improving its core product, ChatGPT. The decision, according to The Wall Street Journal, stems from an internal memo from CEO Sam Altman announcing a “Code Red” mode for the company.

OpenAI is suspending work on side projects and focusing its resources on improving its key product, ChatGPT.

“Code Red” is an internal high alert or emergency priority mode that companies put in place when they believe their key product or business line is at serious risk.

What needs improvement

• personalization;
• speed of work;
• stability and reliability.

Gemini steps up pressure on OpenAI

ChatGPT, which debuted on Dec. 1, 2022, was effectively the growth point for the entire LLM market. However, in recent years, rivals have significantly strengthened their positions.

The biggest challenge for OpenAI is Google, which released a new version of Gemini 3 in early November and has integrated it into most of its products.

Also, late last month Anthropic introduced Claude Opus 4.5, which is actively gaining share among enterprise customers. Alternatives to Meta* (LLaMA) and China’s DeepSeek are also on the market.

Anthropic’s Claude Opus 4.5 has been actively gaining share among enterprise customers.

Growing competition makes it especially important for OpenAI to maintain its leadership – especially as the company continues to operate in a high-cost environment.

Financial pressures and long-term investments

Microsoft, which has invested more than $13 billion in OpenAI and owns about 27% of OpenAI PBC’s commercial entity, reported a $3.1 billion loss in its fiscal first quarter through the end of October related to OpenAI’s operations. That amount had previously been hidden in the “other expenses” line.

At the end of October, the company reported a $3.1 billion loss related to OpenAI’s operations.

Despite no profits in more than a decade, OpenAI has been ramping up its infrastructure investments: building large data centers and making long-term purchases of computing capacity. In October, the company entered into an agreement with AMD to deliver 6 GW of capacity in exchange for a 10% stake in AMD.

And it’s also making a deal with AMD to provide 6 GW of capacity in exchange for a 10% stake in AMD.

The spending raises questions among analysts about how OpenAI will recoup its investment.

Waiting for a new release and the problems of GPT-5

Despite the pressure, OpenAI maintains leadership in a number of areas. According to the WSJ, the company is preparing to release a major new AI model as early as next week — it’s supposed to be a response to the latest Gemini update.

The previous major release, GPT-5, which came out in August 2025, received mixed reviews. Users complained about the “too impersonal” nature of the model and a decline in performance in math and geography. About three months later, OpenAI updated the model and fixed some of the problems.

* Owned by Meta, it is recognized as an extremist organization in Russia and its activities are banned.