OpenAI CEO Sam Altman sent his employees an internal memo that was leaked to the media. In it, he openly acknowledges that Google has recently been doing excellent work across the board, especially in the pretraining of artificial intelligence models. According to him, this could cause temporary economic difficulties for OpenAI. Altman writes that although the company is catching up quickly, the situation is challenging. The memo was written before the launch of the new Google Gemini 3 model, which has been described as Google's most advanced AI system. The model delivers strong results in automated website creation, product design, and programming—areas that generate significant revenue for OpenAI.
In the memo, Altman describes the atmosphere at OpenAI as "rough vibes" and speaks of the economic headwinds facing the company. He acknowledges that revenue growth could fall to single digits by 2026. This stands in stark contrast to his public ambitions, in which he talks about trillions of dollars in investment. According to Altman, competitors such as Google and Anthropic are closing in on OpenAI in the race for the best AI technologies. For example, Google recently took the AI performance crown with its Gemini 3 Pro model.
OpenAI's Slowing Growth
In the memo, Altman warns that OpenAI must remain focused despite short-term pressure from competitors. He emphasizes that most of the research team should remain focused on achieving superintelligence, which he considers crucial. He acknowledges that the company is now "catching up quickly" and needs to focus on highly ambitious bets, even if that means temporarily falling behind in the current paradigm. Altman writes: "We need to stay focused through short-term competitive pressure."
Questions about OpenAI's slowing momentum had already emerged this summer. At a quarterly company earnings meeting, Chief Financial Officer Sarah Friar acknowledged that user engagement with the popular ChatGPT chatbot had weakened, even though overall financial results remained solid. Altman adds in the memo that OpenAI is building the strength to weather great models from competitors and urges employees to remain confident and focused on major goals.
OpenAI is reportedly working on a new language model code-named Shallotpeat. The model is intended to fix errors that arose during the pretraining process. A person familiar with the matter said that the goal is to improve performance and reliability. Altman concludes the memo by saying that despite the company's enormous valuation, "we know we have some work ahead of us, but we are catching up quickly."
Conclusion of the Memo
Altman acknowledges that it is unpleasant for OpenAI to have to do so many difficult things at once. He writes: "It sucks that we have to do so many hard things at once—be the best research lab, have the best AI infrastructure, and be the best AI platform/product company—but that is our lot in life. And I wouldn't trade places with any other company."
OpenAI recently entered into a partnership with Apple's largest supplier, Foxconn, to design and manufacture components for AI data centers in the U.S. Under the agreement, OpenAI will contribute to the development of next-generation server systems and plans to purchase hardware from Foxconn in the near future. Altman described it as a step toward strengthening the American manufacturing base and said that expanding AI infrastructure represents a generational opportunity to reindustrialize America. These partnerships are intended to ensure that the key technologies of the AI era are built in the U.S.
Altman ends the memo with encouragement: "We have built enough strength as a company to weather great models delivered by others... which is why it is critically important that most of our research team remain focused on truly achieving superintelligence." In doing so, he underscores that despite challenges from Google and others, OpenAI remains determined to lead in artificial intelligence.
Sources: theinformation.com, theinformation.com and economictimes.indiatimes.com
