OpenAI Launches GPT-5.2 as It Navigates 'Code Red' (www.wired.com)

OpenAI has launched GPT-5.2, its most advanced AI model to date, achieving significant improvements in writing, coding, and reasoning tasks. This release follows a company-wide "code red" initiated by CEO Sam Altman in response to growing competitive pressure, particularly from rivals like Google's successful Gemini 3 model. While OpenAI asserts that the development of GPT-5.2 was not accelerated by this urgency, it emphasizes a renewed focus on enhancing ChatGPT's capabilities amidst shifting market dynamics. GPT-5.2 introduces three distinct models—Instant, Thinking, and Pro—each optimized for different tasks, with the Thinking model outperforming human professionals in 70% of evaluated tasks and reducing hallucinations by 38% compared to its predecessor. This model not only excels at complex problem-solving but also strengthens responses to sensitive topics, crucial given previous mental health concerns and usage challenges. As OpenAI aims to bolster user engagement and address recent criticism, the successful rollout of GPT-5.2 represents a pivotal step in maintaining its leading position in the AI landscape while navigating the delicate balance of enhancing user experience amid intense competition.