Sina’s AI morning brief spotlights ChatGPT “adult mode”, youth AI usage in Japan and insider‑trading fears around AI prediction markets

Chinese portal Sina’s AI desk published a morning wrap‑up collating overnight global AI stories, ranging from geopolitics to product launches and safety concerns. One highlight is OpenAI’s plan to roll out an “adult mode” in ChatGPT in early 2026, with Fidji Simo saying the company is testing an age‑prediction model to automatically detect under‑18 users and gate explicit content—essentially building an AI bouncer into the chatbot. The brief also cites a Japanese LINE Yahoo survey showing that 15‑ to 24‑year‑olds most often use AI for casual web queries and homework help, with young women disproportionately turning to chatbots for advice and emotional support, underscoring how quickly AI is becoming a social confidant. Other items include Google’s Gemini‑powered real‑time earbud translation pilot in the U.S., Mexico and India, and a spate of “insider trading” allegations on crypto prediction market Polymarket after traders made million‑dollar wins by betting on the exact timing of OpenAI’s GPT‑5.2 launch and Google’s year‑end trends lists—prompting some tech firms to ban employees from such platforms. Overall, the brief paints a picture of AI seeping into everyday life while regulators and companies scramble to close gaps in content moderation, youth safety and information‑leak risks.