OpenAI has declared a "code red". Google threatens to take away its leadership in the AI race

OpenAI CEO Sam Altman has announced to employees that he's instituting a "code red" mode to urgently improve the quality of chatbot ChatGPT, The Information and The Wall Street Journal reported independently, citing an internal company memo. In order to focus on its key product, OpenAI is forced to postpone the development of other projects, including the launch of advertising tools, the development of AI agents for health and shopping, as well as the digital assistant Pulse, the WSJ wrote.

OpenAI's color-coding system uses "code red" to indicate issues of the highest urgency. In a statement to staff, Altman said the company has a lot of work to do on the usability of the chatbot, including improving personalization features, increasing speed and reliability, and expanding the range of questions it can answer. To address these challenges, the company is introducing daily reporting meetings, as well as the practice of temporarily reassigning employees to teams responsible for ChatGPT development.

OpenAI's move to work in emergency mode is due to growing pressure from competitors. According to WSJ, Altman is particularly concerned about Google, which in November released the third version of its Gemini AI model, which outperformed ChatGPT in benchmark tests and led to a sharp rise in the search giant's stock. Gemini's customer base has been growing since the August release of the Nano Banana image generator - according to Google, the monthly number of active users rose from 450 million in July to 650 million in October. In the enterprise segment, OpenAI is under pressure from AI startup Anthropic.

Despite having more than 800 million users per week and being the leader in the AI race, OpenAI remains unprofitable and has to aggressively raise funding, WSJ notes. This puts the startup in a financially vulnerable position compared to competitors like Google, which is able to fund development out of current revenues.

In recent months, OpenAI has struggled to find a balance between chatbot security and maintaining customer engagement. The GPT-5 model, released in August, drew complaints from some audience members who complained about the "cold tone of responses" and difficulties in solving simple math and geography problems. In November, OpenAI updated the chatbot, making the tone of communication warmer and improving the quality of execution of user instructions. Altman said the new version of ChatGPT, scheduled for release next week, is ahead of Google's latest developments, the WSJ notes.

This article was AI-translated and verified by a human editor