Sam Altman has activated high alert on OpenAI. Just like share From Wall Street Journal, the company’s CEO announced this Monday in an internal memo that the company enters “code red” to improve ChatGPTthe tool that has catapulted the company to stardom but that now sees its rivals closing the gap at breakneck speed.
what’s happening. OpenAI is postponing several important projects to focus all its resources on improving the daily ChatGPT experience, according to the internal memo to which WSJ has had access. According to Altman, the chatbot urgently needs advances in personalization, speed, reliability and the ability to answer a broader range of questions. Among the postponed projects are initiatives to include advertising in the free version of ChatGPT, AI agents for health and purchases (the latter was announced very recently), and Pressa personal assistant in development.
why now. The pressure comes mainly from Google. Your model Gemini 3released last month, has outperformed OpenAI in industry benchmarks and sent the Mountain View giant’s stock soaring. Just like assures In the middle, Gemini’s monthly active users went from 450 million in July to 650 million in October, a meteoric growth that sets off all the alarms at OpenAI. Although ChatGPT maintains the lead with approximately more than 800 million weekly users, the speed at which Google is gaining ground is worrying.
The underlying problem. OpenAI is in a delicate position. The company it is not profitable and it needs constant rounds of financing to survive, which puts it at a disadvantage compared to Google and other technology companies that can finance their investments with their own income. It’s also spending more aggressively than its main startup rival, Anthropic.
According to their own financial projectionsOpenAI will need to reach revenues of approximately $200 billion to be profitable in 2030. All while being committed to investments of hundreds of billions in data centers.
The last setbacks. The company has had a difficult time lately balancing the security of its chatbot with making it more attractive to users. The GPT-5 model Launched in August, it disappointed some users, who complained about its colder tone and problems answering simple math and geography questions. OpenAI had to update the model last month to make it warmer and better able to follow user instructions.
OpenAI’s response. According to point In the middle, Altman has established daily calls for those responsible for improving ChatGPT and has encouraged temporary team transfers. WSJ assures that the company uses three color codes: yellow, orange and red, to describe the different levels of urgency necessary to address problems.
According to the outlet, prior to this “code red”, OpenAI had declared a “code orange” in its effort to improve the chatbot. Nick Turley, Head of ChatGPT at OpenAI, stated in X that ChatGPT represents 70% of global AI-assisted activities and 10% of search activities.
An unexpected script twist. This represents a radical change compared to three years ago, when it was Google who declared its own code red in response to the threat posed by ChatGPT. And after a groundbreaking Google I/O Last May, those from Mountain View have witnessed brutal growth in all the directions in which the AI race is currently pointing, with improvements in their chatbot, the deployment of countless AI agents, improvements in their applications and more. Now it seems that it is OpenAI who must defend its position.
And now what. Altman advertisement that next week OpenAI will launch a new reasoning model that, according to internal evaluations, surpasses Google’s Gemini 3. However, he acknowledges that there is still a lot of work to be done in the everyday chatbot experience.
Cover image | OpenAI and Xataka Android
In Xataka | China already has an army of 5.8 million engineers. His new plan involves accelerating doctorates
GIPHY App Key not set. Please check settings