OpenAI’s CEO Sam Altman on Monday declared a company-wide “code red,” mobilizing the startup to urgently overhaul its flagship product ChatGPT as competitive pressure intensifies, especially from Gemini 3, the latest generative-AI offering from Google. 

An internal memo points to mounting concern that ChatGPT, once seen as the unchallenged leader in conversational AI, is falling behind. Gemini 3 has recently topped benchmark tests, putting Google’s model on track to erode OpenAI’s head start. 

Altman warned employees that the competition, particularly from Google, risks creating economic headwinds. The message recognizes a narrowing gap as rivals like Google and Anthropic ramp up their AI efforts.

OpenAI will delay other products and initiatives, including earlier plans for advertising, and AI agents for shopping, health, and a personalized assistant called “Pulse.”

The firm will re-prioritize development of ChatGPT,  focusing heavily on improving speed, reliability, personalization, and model behavior, including its image-generation capabilities and addressing issues like “overrefusals” (when the AI declines to answer benign questions).

OpenAI is ramping up what it called a “surge,” with the goal of shipping a new reasoning model next week that — according to internal evaluation — already outperforms Gemini 3.