OpenAI announced the launch of its new GPT-5.2 model just one month after its last ChatGPT update, amid fierce competition with Google's Gemini 3 model. The company markets the new model as "the best release for daily professional use." GPT-5.2 will support the ChatGPT platform and will be available to developers via an API, with its rollout beginning today for subscribers to paid plans. The company confirms that the new version achieves qualitative leaps in real-world applications and evaluation standards in programming, mathematics, science, vision, and logical reasoning. It also outperforms the previous version in handling long-context queries and is less likely to provide incorrect information. While OpenAI faces lawsuits and sharp criticism for rushing its models to market despite security risks, Fidji Simo, the head of applications, denied that the release was rushed, stating that GPT-5.2 is safer than previous versions. She stated: "We are improving in almost every dimension of safety, whether that's self-harm or different types of mental health problems or emotional dependence." The company also revealed the delay of the "adult mode" launch to the first quarter of next year, after announcing it would be released in December 2025, noting that it has begun testing an age-gating system in some countries. Simo said: "We designed GPT-5.2 to unlock greater economic value for people," adding that it is "better at creating spreadsheets, building presentations, writing code, perceiving images, understanding long context, using tools, and processing complex, multi-step projects." As the need to monetize ChatGPT to raise the trillions needed to support OpenAI grows, observers are wondering when advertising will appear in the chatbot, but Simo confirmed that "there is nothing to announce on ads today," adding that if the company ever decides to enable ads, it will do so "in a way that respects the very close relationships that people have with ChatGPT".
