Sam Altman Warns OpenAI Staff of ‘Economic Headwinds’ Amid Google’s Gemini 3 Comeback

The boss of the company that started the AI craze is telling his team to buckle up for a bumpy ride.

OpenAI CEO Sam Altman recently dropped a startling memo to his staff, admitting that the company is bracing for a tough competitive spell, or as he put it, “rough vibes.” The reason? Google’s major comeback in the AI race, fueled by its deep pockets and the widespread rollout of its newest model, Gemini 3.

The internal memo, reportedly leaked to The Information, signals a significant shift in the competitive landscape that OpenAI’s ChatGPT has long dominated. Altman acknowledged that “Google has been doing excellent work recently in every aspect,” noting that Google’s progress in areas like pre-training could affect OpenAI’s position.

He also warned employees that this resurgence could “create some temporary economic headwinds for our company.” While OpenAI is still highly valued, analysts are noting the pressure on it. Google, a “full-stack company,” holds a massive economic advantage, controlling everything from its own hardware (TPU development) to its vast user platforms, including Search, Workspace, and Android.

The release of Gemini 3 has only intensified the challenge. Google has been integrating this powerful AI directly into its ecosystem, providing billions of users with access to its new capabilities, often without them even realizing it. This ubiquity creates a powerful network effect that standalone companies like OpenAI, which rely on users intentionally seeking out ChatGPT, find hard to counter.

The company’s CFO, Sarah Friar, has also recently informed investors that engagement with ChatGPT has softened, even as overall financials remain stable.

Catching up and superintelligence

Despite the challenging forecast, Altman maintained a confident stance about OpenAI’s future. He acknowledged that rivals like Google and Anthropic are closing the gap, but assured staff that OpenAI is “catching up fast.”

The CEO emphasized the need for his team to focus on the company’s most ambitious goal, achieving superintelligence.

“We have built enough strength as a company to weather great models shipping elsewhere competition… (so), having most of our research team focused on really getting to superintelligence is critically important,” Altman reportedly wrote in the internal memo.

He also acknowledged the immense challenge of OpenAI’s many roles, stating, “It s**ks that we have to do so many hard things at the same time — the best research lab, the best AI infrastructure company, and the best AI platform/product company — but such is our lot in life. And I wouldn’t trade positions with any other company.”

Strategic response and future focus

The urgency in Altman’s message stems from Google’s technological leaps, as noted by outside experts. For example, the Gemini 3 Pro model has reportedly achieved higher scores on some benchmarks, such as the LM Arena leaderboard, than OpenAI’s current models.

In response, OpenAI is reportedly working on a new language model codenamed “Shallotpeat,” specifically intended to address issues in the pre-training process.

Altman’s core message to his team is to stay focused through this “short-term competitive pressure” and “very ambitious bets,” even if it means getting “temporarily behind in the current regime.” 

This memo shines a light on the new, more competitive phase of the AI industry. The initial shock and awe of ChatGPT’s debut have worn off, the unchallenged era of OpenAI’s dominance is over, and deep-pocketed giants like Google are now fully in the game. For OpenAI, the path forward involves navigating these “rough vibes” while proving it can still innovate faster than anyone else.

Elsewhere in the AI race, a new look at Gemini 3, Google’s most powerful AI model yet, explores how advanced reasoning, interactive Search features, and autonomous coding tools are rolling out together.