Here is a short, human-written style rewrite of the news with only the main important points included:
OpenAI has released its newest AI models — GPT-5.2 Pro and GPT-5.2 Thinking — claiming major improvements in math, science, and reasoning. The company says these upgrades bring it closer to achieving artificial general intelligence (AGI).
The launch comes as competition with Google intensifies. Google’s Gemini 3 Pro recently outperformed OpenAI’s previous GPT-5.1 in several reasoning and multimodal tests. Early users say GPT-5.2 is noticeably more advanced, but it’s still uncertain whether it will surpass Gemini 3.
Despite concerns about high infrastructure costs, CEO Sam Altman said OpenAI remains confident it can grow revenue to match its massive investments.
OpenAI also confirmed that a new ChatGPT “adult mode” is expected next year, though age-verification improvements are still needed. This follows lawsuits accusing the company of allowing harmful interactions with minors.
OpenAI executives acknowledged a “red alert” internally due to Google’s rapid progress but denied rushing their model releases.
