News

OpenAI has declared 'code red' multiple times, most recently in response to Google Gemini benchmark gains

Dec 19, 2025

Key Points

  • OpenAI triggered code red status after Google's Gemini model outperformed its offerings on multiple benchmarks, signaling material competitive threat.
  • Sam Altman directed staff to reallocate resources toward accelerating ChatGPT improvements while shelving work on AI agents and advertising products.
  • OpenAI is prioritizing near-term competitive positioning over platform expansion into new verticals.

Summary

OpenAI has declared code red multiple times in recent weeks. The most recent trigger was Google's release of a new AI model that outperformed OpenAI's best offerings on several benchmarks. Two weeks after Google's launch, Sam Altman directed staff to reallocate resources toward accelerating ChatGPT improvements while delaying work on autonomous AI agents and advertising products. The move reflects competitive pressure from Google's Gemini, which demonstrated benchmark gains that OpenAI views as a material threat. By consolidating resources onto core product performance rather than new verticals, Altman is prioritizing near-term competitive positioning over broader platform expansion.