Gone are the days when OpenAI outrageously dominated the generative AI sector, both in benchmarks and in minds. Today the company is under pressure from its competitors, in particular Google, and Sam Altman is worried.
The battle is underway at OpenAI. In a leaked internal memo, Sam Altman raises concerns about the performance of Google’s competing Gemini 3 Pro model, saying the company has “done a great job” and predicting tough times ahead as competition for cutting-edge AI models accelerates.
Especially since the release of Gemini 3 coincides with another dangerous announcement for OpenAI. On November 18, Microsoft and Nvidia, two of OpenAI’s main investors, announced the investment of $15 billion in Anthropic, a rival start-up to OpenAI, founded by former members of the company, including Dario Amodei, the current boss of Anthropic. The valuation of the young growth is now estimated at 350 billion dollars, compared to 500 billion for OpenAI.
OpenAI loses its technological advantage
Of course, competition over the most advanced AI models is not new. “The cutting-edge AI environment is very competitive, with a handful of players going neck and neck, particularly OpenAI, Anthropic, xAI and Google. That’s not new,” notes Antoine Chkaiban, consultant at New Street Research, a market intelligence firm. On the other hand, having launched ChatGPT before everyone else, OpenAI benefited for a long time from a head start, due to the fact that it was the first company to position itself on this generative AI market. The release of Gemini 3 confirms the fact that this advantage has now evaporated.
“Eighteen months ago, OpenAI was the undisputed king, while Google was seen as a bureaucratic dinosaur, incapable of launching a product to save its skin. The “Code Red” triggered by the company created skeptics, as did Bard’s failed demonstration. Google was already seen as the next IBM — a giant of the past doomed to be devoured by the cool and dynamic young people of OpenAI. But the giant has woken up, and it is winning,” says Derik David, an independent analyst specializing in digital technologies.
This means that Google now has more cards in hand to put pressure on its competitor. “OpenAI has lost its technological advantage, Google has caught up. What will happen if Google decides to cut prices to compete with OpenAI?” asks Gary Marcus, American AI expert.
Added to this are worrying signs for OpenAI in the use of its products. User engagement is stagnating, as CFO Sarah Friar recently admitted. In his internal note, Sam Altman anticipates a very strong slowdown in the company’s turnover growth which, according to him, could be around 5 to 10% next year, due in particular to competition from Claude (Anthropic) and Gemini (Google). We are far from the 300% estimated for 2025.
To regain the initiative, the company is currently working on the Shallotpeat project, a code name hiding a model aimed at resolving bugs in the training of AI models.
A matter of ecosystem
The fact that Google has caught up technologically with OpenAI is all the more annoying for the latter as it does not have the powerful and extensive ecosystem that the Californian giant benefits from thanks to its decades of seniority and its solid hold on the web. “Google has the world’s data, makes its own chips, and has unlimited cash. Google also has YouYube, its search engine, Gmail, Maps, and Android, as well as billions of users. No other company has such a combo,” notes Andrew Lokenauth, author of The Finance Newsletter, a newsletter for investors.
This ecosystem gives Google a significant advantage in its fight with OpenAI, illustrated during the launch of Gemini 3: as soon as it was deployed, it was immediately integrated into the Google search engine, AI Mode, Google Workspace and Android, immediately reaching billions of users without them needing to download anything.
However, the battle for AI is not only about technological superiority, but also and above all about the ability to bring value to companies, by integrating this technology into a coherent ecosystem. “Purely technical innovations between cutting-edge models are less and less a distinguishing factor. What will make the difference is above all the ability to bring value to companies, to understand the structure of their data, to make it usable, and to integrate their workloads in a fluid manner. Microsoft is very well positioned on this. This is why Azure is showing such growth and is beating AWS on the cloud dedicated to AI,” notes Antoine Chkaiban. This is what a company like Palantir has understood, which aims to place itself at the heart of the transformation of businesses through AI, and to do so has garnered the confidence of investors.
OpenAI, for its part, has its historic partnership with Microsoft. This is why the merger of the latter with Anthropic should cause concern for Sam Altman’s company. Certainly, it voluntarily chose to distance itself from the IT giant (which nevertheless remains one of OpenAI’s main investors), in order to plan its IPO and to have more room to maneuver to forge agreements beyond the Microsoft ecosystem. A bet which is however risky, since it encourages Microsoft to find alternatives to OpenAI to put AI at the heart of its ecosystem.
Finally, to all this are added the remaining fears around a bursting of the AI bubble, which Nvidia’s latest good results have not been enough to allay, and which would put OpenAI, which is fueling its crazy growth by piling up debt, in a very delicate situation.
