
Sam Altman pauses OpenAI projects to strengthen ChatGPT
OpenAI has declared an internal “code red,” accelerating the release of its next major AI model, GPT-5.2, a direct response to rising competition from Google’s Gemini 3. The update, originally slated for later in December, is now expected to arrive as soon as December 9, 2025.
The push comes after Gemini 3 stunned the AI world with strong benchmark results, prompting concern at OpenAI about losing its edge. In response, company teams have paused side-projects and shifted full focus onto enhancing ChatGPT’s core performance.
Unlike past updates that added flashy new features, GPT-5.2 aims for under-the-hood improvements: faster responses, more accurate reasoning, improved reliability, and better customization. Internal testing reportedly shows GPT-5.2 outperforming Gemini 3 in reasoning tasks, though independent verification is still pending.
The release date could shift if technical or deployment issues arise; OpenAI seems to be planning a gradual rollout, prioritizing Plus and Enterprise users first before wider distribution.
Overall, GPT-5.2 reflects a strategic pivot for OpenAI. With competition accelerating, the company is doubling down on reliability and performance rather than feature-bloat. If all goes as planned, this update could help restore confidence among users and maintain OpenAI’s leading position in the fast-evolving AI landscape.