'Code Red' says OpenAI CEO Sam Altman in an internal memo post Google Gemini 3 release: Report

OpenAI Chief Executive Sam Altman has reportedly declared a “code red” at the company, demanding immediate and focused efforts to significantly improve the capabilities of the company's flagship product, ChatGPT, The Information reported. This urgent directive underscores the intensifying competition within the sector, primarily driven by the launch of Google’s Gemini 3 AI model.

In an internal memo to staff, Altman stressed that the company is at a “critical time” for ChatGPT. The urgency stems from the impressive performance of Google’s latest offering, which has reportedly outperformed rivals on several key industry benchmarks. Altman cautioned employees that the market reaction to Gemini 3 could lead to “temporary economic headwinds” for the San Francisco-based firm, acknowledging that the “vibes out there [will] be rough for a bit.”

The competitive threat was amplified when prominent figures publicly switched allegiance to the Google product. Marc Benioff, the chief executive of software giant Salesforce, wrote that he was “not going back” after testing Gemini 3. Benioff described the generational leap in its reasoning, speed, images, and video handling as “insane,” suggesting the release has fundamentally changed the landscape once again.

Despite the pressure, ChatGPT remains a globally dominant tool with approximately 800 million weekly users. Nevertheless, OpenAI is taking corrective measures. The firm is now diverting more internal resources toward product enhancement and has opted to delay a planned foray into introducing advertising within the chatbot to maintain focus on core improvements.

Financially, OpenAI is operating at a loss but anticipates annual revenues of over £16 billion (around $20 billion) this year, with Altman projecting growth into the hundreds of billions by 2030. The company's latest valuation reached a substantial £400 billion. However, the long-term strategy is exceptionally capital-intensive, requiring a pledge to spend $1.4 trillion on data centre costs over the next eight years to train and operate its systems. The “code red” thus reflects both a product crisis and a necessity to secure a dominant market position to justify these vast future expenditure commitments.

“Based on the trends we are seeing of how people are using AI and how much of it they would like to use, we believe the risk of OpenAI of not having enough computing power is more significant and more likely than the risk of having too much,” said Altman last month.