According to foreign media reports, in early December 2025, OpenAI CEO Sam Altman said in an internal memo and Slack message that he was activating a "code red" (red alert) for ChatGPT, saying that "this is a critical period for ChatGPT." Against the backdrop of slowing user growth and increased competition from Google Gemini, he decided to deploy more employees to improve the product, focusing on serving more than 800 million weekly active users.
According to the memo, current priorities include: enhancing personalization capabilities that allow users to customize chat styles and behaviors; Improve the quality of ImageGen image generation; Optimize model behavior, responsiveness, and stability, and minimize unnecessary rejections. In contrast, OpenAI will postpone the pace of a series of projects, including advertising business tests that have not yet been publicly acknowledged, AI agent products for shopping and health scenarios, and personalized information flow products such as Pulse. The company also plans to release a new inference model next week, with internal testing claiming that it outperforms Google's Gemini 3 in inference performance.
The industry generally believes that this adjustment means that OpenAI has chosen to put the core product experience of "protecting the disk" before new business expansion in the short term under the pressure of high-investment infrastructure and fierce market competition, and strive to stabilize user stickiness and brand reputation by improving the personalization, speed and reliability of ChatGPT.
Frequently Asked Questions
Q: What exactly does OpenAI's internal "code red" refer to?
A: It mainly refers to increasing priorities within the company, focusing more human resources on improving the ChatGPT experience, and suspending the progress of some new projects.
Q: Why is OpenAI activating a "red alert" now?
A: The report mentioned that on the one hand, the growth of ChatGPT users has slowed down, and on the other hand, Google Gemini 3 has put more competitive pressure on OpenAI in benchmarking and product promotion.
Q: What aspects of ChatGPT will be focused on improving this time?
A: This includes stronger personalization, improved image generation performance, optimized model behavior, faster response speed, and improved service reliability, while reducing excessive rejections.
Q: Which OpenAI projects will be delayed?
A: Including advertising plans that are still in beta, AI agent products for shopping and health, and Pulse, a personalized messaging service, have been given way to ChatGPT experience optimization.
Q: What is the situation with the new reasoning model mentioned?
A: According to internal information, OpenAI plans to release a new reasoning model next week, and internal tests believe it outperforms Google Gemini 3, but specific technical details and public test results have not been disclosed.
