OpenAI Expands Its Reach: Shopping, Social Video, and Safety Tools for Teens

Turning Chat Into a Checkout

OpenAI has begun weaving commerce directly into ChatGPT^[1], a move that positions the chatbot not only as a source of information but also as a new point of sale. U.S. users across free and paid tiers can now buy products from Etsy sellers without leaving a conversation, with access to more than a million Shopify merchants expected in the near future.

The Instant Checkout system allows someone to ask for gift suggestions, browse relevant options surfaced in the chat, and then complete the purchase with a stored card, Apple Pay, or other payment services. The transaction itself runs through the merchant’s own systems, while ChatGPT acts as the go-between, passing on the necessary details securely. Shoppers see the same prices as they would on a merchant’s site, but the seller pays a small transaction fee.

At the core of this feature is the Agentic Commerce Protocol, a standard OpenAI developed alongside Stripe to help AI systems and businesses complete orders together. The company has open sourced the protocol to encourage adoption, offering developers and retailers a straightforward way to connect their systems to agentic shopping flows. While the first release supports single-item transactions, OpenAI says it is working on multi-item carts and regional expansion.

This move pushes the company into direct competition with Google and Amazon, which have long shaped how online retail operates. Search results and marketplace algorithms have historically dictated what products reach customers first. If more people begin to shop through AI conversations, the balance of influence could shift toward the developers of these agents, who then decide what results appear and how fees are structured.

Building a Social Video Platform Around Sora 2

Alongside its e-commerce efforts, OpenAI is preparing^[2] to enter the social media space with a new app powered by its Sora 2 video model. Reports suggest the software will resemble TikTok in format, offering a vertical feed of short videos navigated with swipes. What sets it apart is that every clip will be AI-generated rather than uploaded from a user’s camera roll.

The first version of the app is expected to limit clips to around ten seconds, shorter than the typical uploads supported by TikTok. An identity verification tool is also part of the design. If users opt in, the model can generate content featuring their likeness, allowing others to remix their digital persona into different videos. Whenever someone’s likeness is used, a notification will be sent, even if the clip never makes it to the public feed.

To address rights concerns, the system is built to block certain copyrighted materials, although early reports indicate that enforcement may rely on rights holders opting out. The model itself will refuse some prompts entirely, reflecting the growing attention on copyright in generative media.

By focusing on an AI-only feed, OpenAI is testing whether audiences will consume entertainment that has no direct human authorship. The experiment also expands the reach of Sora beyond experimental clips into a social context, creating a new setting for AI-generated culture.

Introducing Parental Controls for ChatGPT

While building out new products for commerce and media, OpenAI is also responding to pressure over safety. The company has rolled out parental controls for ChatGPT^[3] across the web, with mobile access coming soon. Parents can now link accounts with their teenagers to set restrictions and monitor use more closely.

The controls cover several areas. Parents can reduce or block access to sexual roleplay, violent scenarios, extreme beauty ideals, and other sensitive themes. They can also disable voice mode, image generation, or the system’s ability to remember past chats. Turning off memory reduces personalization but may strengthen guardrails by preventing conversations from gradually drifting into unsafe territory.

Another option allows parents to stop transcripts from being used to improve OpenAI’s models, giving families greater control over how their data is handled. Quiet hours can be scheduled so that teenagers cannot use ChatGPT during set times of the day. Notifications are also available if the system detects signs of serious safety risks, with alerts sent by email, text message, or app push.

Parents do not gain access to their children’s chat history. The link instead provides settings and alerts, with limited exceptions if urgent risks are identified. Teenagers retain some autonomy as well, since they can disconnect the link, though parents are notified when this happens.

The changes follow months of scrutiny. After a teenager in the U.S. died by suicide earlier this year, with allegations that the chatbot had played a role, lawmakers and grieving families called for stronger protections. OpenAI has since been working on an age-estimation system to better identify underage users and apply safeguards.

A Broader Push Into Consumer Life

These three developments — commerce integration, an AI-driven video app, and parental safety controls — point to OpenAI expanding beyond its roots as a research company into a platform that touches daily life in different ways. Each move carries implications not just for users but also for competitors, regulators, and industries built on existing digital habits.

The shift into shopping challenges the long-standing dominance of search engines and marketplaces in directing retail. The video app tests whether AI can generate an entertainment ecosystem compelling enough to rival human creators. And the safety measures show that the firm cannot avoid the responsibilities that come with shaping how young people interact with intelligent systems.

Together, these initiatives sketch the outline of a company positioning itself at the center of online discovery, culture, and trust. The path forward will depend on how widely these tools are adopted, how well they are managed, and how regulators respond to the risks they introduce. For now, they mark a significant step in OpenAI’s transformation from a lab building models into a consumer-facing force in technology.

Note: This post was edited/created using GenAI tools.

Read next:

• Generative AI Becomes Two-Way Force, Altering Company Marketing and Consumer Product Searches^[4]

• Americans Pull Back on Subscriptions as Costs Rise and Habits Shift^[5]