OpenAI has outlined a series of strategic changes across its AI ecosystem, focusing on safety, product direction, and platform consolidation. The company introduced a new set of prompt-based safety policies designed to help developers build safer AI experiences for teenagers, working in collaboration with Common Sense Media and everyone.ai. The framework addresses risks including harmful content, dangerous behavior, and age-restricted interactions, and is designed to be compatible across models, including its gpt-oss-safeguard system.
In parallel, OpenAI is restructuring its approach to AI-driven commerce within ChatGPT. The company confirmed it is deprioritizing its “Instant Checkout” feature after limited user adoption, shifting instead toward positioning ChatGPT as a centralized product discovery and research tool. This strategy is supported by its Agentic Commerce Protocol, developed with Stripe, enabling merchants to manage their own checkout experiences while ChatGPT focuses on surfacing product information.
Separately, OpenAI is discontinuing its Sora social app just six months after launch. The TikTok-style platform struggled to sustain engagement and faced ongoing moderation challenges related to AI-generated content. While the app is being shut down, OpenAI confirmed that its underlying Sora 2 video-generation model will remain accessible within ChatGPT.
