OpenAI Launches GPT-5.2 Amid “Code Red” Push and Growing Competition

OpenAI Launches GPT-5.2 Amid “Code Red” Push and Growing Competition

Key Points

  • OpenAI launches GPT-5.2 as part of a “code red” effort to boost ChatGPT.
  • The model family includes Instant, Thinking and Pro tiers for different use cases.
  • GPT-5.2 scores highest on the GDPval benchmark, beating humans on over 70% of tasks.
  • Hallucinations are reduced by 38% compared with GPT-5.1.
  • Competition from Google’s Gemini 3 intensifies pressure on OpenAI.
  • New safeguards address mental‑health‑sensitive prompts and age‑based content protection.
  • An “adult mode” is planned for early 2026.
  • OpenAI targets a 5% increase in daily active users before 2026.

OpenAI Launches GPT-5.2 as It Navigates ‘Code Red’
Image may contain Electronics Mobile Phone Phone Silhouette Adult Person and Mat

Image may contain Electronics Mobile Phone Phone Silhouette Adult Person and Mat

OpenAI’s New Model Release

OpenAI unveiled GPT-5.2, describing it as the company’s best model for everyday professional use. The launch comes under a “code red” initiative that signals a focused allocation of resources to improve ChatGPT amid what executives called “the greatest competitive pressure we’ve ever seen.” The GPT-5.2 family is divided into three tiers: Instant, which prioritizes speed and information‑finding; Thinking, optimized for coding, math and planning; and Pro, the most powerful version delivering higher accuracy on difficult questions.

Performance Benchmarks

According to OpenAI, GPT-5.2 achieved the highest scores to date on the GDPval benchmark, a test that compares AI performance to human professionals across 44 real‑world occupations. The model outperformed humans on more than 70 percent of tasks and completed them 11 times faster. In addition, the company reported a 38 percent reduction in hallucinations compared with the previous GPT-5.1 model on factual‑question benchmarks.

Competitive Landscape

The release occurs as Google’s Gemini 3 model gains traction, boasting over 650 million monthly active users, while OpenAI’s ChatGPT reports roughly 800 million weekly active users. This rivalry has prompted OpenAI to pause some ambitious projects, such as introducing ads to ChatGPT, and instead concentrate on core technology improvements.

Safety and Mental‑Health Features

OpenAI highlighted enhancements to how ChatGPT handles sensitive prompts related to self‑harm, mental‑health distress, or emotional reliance on the model. The firm also mentioned early steps toward an age‑prediction system that would automatically apply content protections for users estimated to be under 18. An “adult mode” allowing users over 18 to engage in erotic conversations is slated for rollout in the first quarter of 2026.

User Growth Goals

Internally, OpenAI set a target to increase daily active users by 5 percent before 2026, reflecting the pressure to expand its user base while maintaining a balanced conversational experience that avoids excessive agreeableness.

Outlook

With GPT-5.2, OpenAI aims to reinforce ChatGPT’s performance, safety and appeal as it navigates a intensifying AI market. The company plans to make the new models available to both ChatGPT users and developers via its API, emphasizing clear gains across everyday and advanced use cases.

Source: wired.com