Sam Altman (ChatGPT) sounds the alarm to OpenAI’s internal teams about the rise of Google’s Gemini

There is a wave of panic in the corridors of OpenAI. The American company, famous for its ChatGPT language model and its Sora image and video generator, saw its dominance in the consumer AI agent market undermined by the thunderous arrival of the Gemini replica by Google. According to an article in Wall Street Journalthe founder of OpenAI, Sam Altman, would have sent a note to the teams to declare the “code red“, or the highest level of warning used internally, on the quality of ChatGPT, thus taking first priority ahead of all other initiatives planned in the short and medium term.

Thus, the launch of the final version of the search assistant Pulse or the integration of advertising models within the GPT offer is for the moment put aside in favor of the development effort on ChatGPT. The objective for OpenAI is to make the LLM more “staff“, but also more responsive and versatile in the number of questions it can answer. Nothing is mentioned to reinforce security for the most vulnerable users – last November 6seven individuals filed a complaint in California against OpenAI for involuntary manslaughter and assisted suicide after the deaths of 4 GPT-4o users.

Since its launch in version 3.0, Gemini – and especially its video generator via text prompts Nano Banana – displays ever-growing popularity, with 650 million monthly active users – 300 million more than last March. Alphabet, Google’s parent company, is catching up with OpenAI and ChatGPT’s 800 million weekly users. The pausing of advertising projects and secondary agents could, however, undermine the financial objectives of OpenAI which planned to generate 20% of revenue from these two new products. The firm announced a net loss of $13.5 billion for $4.3 billion over the first six months of 2025.