Key takeaways OpenAI launched Shopping Research, an AI-powered tool that creates personalized buyer's guides for ChatGPT users through reinforcement learning. The feature is powered by GPT-5 mini and...
OpenAI announced Monday the launch of Shopping Research, a specialized tool designed to help ChatGPT users navigate complex purchasing decisions through AI-powered product discovery and comparison.
The feature arrives just ahead of the holiday shopping season and represents OpenAI's latest move into the competitive ecommerce space.
The Shopping Research feature transforms product searches into conversational experiences.
[caption id="attachment_32975" align="aligncenter" width="800"]
Image Source: OpenAI[/caption]
Users can describe what they need, such as "Find the quietest cordless stick vacuum for a small apartment" or "I need a gift for my four-year-old niece who loves art," and the system generates an interactive quiz to clarify priorities like budget, features, and preferences.
Unlike standard ChatGPT responses that take seconds, Shopping Research operates over several minutes, scanning numerous websites for pricing, availability, reviews, specifications, and images.
Users can guide the research in real time by marking products as "Not interested" or "More like this," allowing the AI to refine its recommendations dynamically.
The tool delivers personalized buyer's guides highlighting top products, key differences between options, and trade-offs based on user constraints.
It draws on ChatGPT's memory feature to incorporate information from past conversations, such as user interests or preferences previously mentioned.
"[Shopping research] will actually look at dozens and dozens of websites and sources, and it will really index on all the implicit and explicit requirements that you're giving it," said Manuka Stratta, a member of OpenAI's technical staff, in remarks to reporters at a press event.
Shopping Research is powered by a version of GPT-5 mini that has been trained with reinforcement learning specifically for shopping tasks.
According to OpenAI, the specialized model reads trusted sites, cites reliable sources, and synthesizes information across multiple sources to produce high-quality product research.
The company reports significant accuracy improvements with the new tool.
OpenAI's latest model accurately identifies items matching all user criteria 64% of the time, a substantial increase from the 37% success rate observed in previous ChatGPT iterations.
The system prioritizes reviews from sources OpenAI considers higher-quality, such as user experiences shared on Reddit, over paid marketing or product page reviews.
"Generally, a lot of reviews on Reddit are pretty trustworthy," Stratta told reporters ahead of the announcement.
Isa Fulford, who leads OpenAI's Deep Research and ChatGPT agent teams, acknowledged the challenges in distinguishing authentic reviews.
"It's impossible to get it 100% correct," she said, noting that identifying legitimate reviews is "a pretty hard task."
The tool is designed to avoid low-quality or spammy websites, defined as sites with excessive pop-up advertisements or clearly promotional content.
However, it will respect user requests for specific retailers, including discount platforms like Temu if explicitly asked.
Shopping Research began rolling out Monday to all logged-in ChatGPT users globally across Free, Go, Plus, and Pro plans. OpenAI is providing nearly unlimited usage through the holidays to assist with gift shopping and product inspiration.
For ChatGPT Pro subscribers, the feature integrates with Pulse, which can proactively suggest personalized buyer's guides based on previous conversations.
One significant limitation involves Amazon, the largest ecommerce platform in the United States. OpenAI confirmed the tool only accesses websites that permit its browsing agents, respecting each site's robots.txt file.
"We respect all of the OpenAI robots.txt," Fulford told reporters. "Anything that allows us to access their site, our product will access, and anything that we don't, we won't access."
Amazon has updated its robots.txt file to block several OpenAI crawlers, including those used for model training, real-time browsing, and its SearchGPT engine.
This means Amazon product listings may be limited or entirely absent from ChatGPT's Shopping Research results.
Stratta told reporters she views Amazon's AI shopping assistant Rufus and ChatGPT Shopping Research as "complementary" tools rather than direct competitors.
Despite partnerships with retailers like Walmart, Etsy, and Shopify merchants, Fulford emphasized that Shopping Research does not prioritize partner sites.
"We don't have any prioritization for any specific sites," she said, adding that the tool focuses on finding strong matches for user inquiries.
OpenAI has been expanding its presence in ecommerce through multiple initiatives.
The company introduced Instant Checkout in September, which allows users to purchase products directly from eligible merchants through ChatGPT using the Agentic Commerce Protocol built with Stripe.
Shopping Research is currently separate from Instant Checkout functionality, requiring users to click through to retailer websites to complete purchases.
However, OpenAI confirmed in its blog post that it plans to integrate the two features in the future, though no timeline was provided.
The launch intensifies competition in AI-powered shopping as OpenAI, Google, Amazon, and startups like Perplexity AI vie to become integral to consumers' purchasing processes.
Amazon reports that shoppers using its Rufus assistant are 60% more likely to complete purchases, with the tool expected to generate over $10 billion in annual sales.
"Already, hundreds of millions of people ask ChatGPT for product advice, and we see this especially in more spec-heavy categories like electronics, health, beauty, and home," Fulford said, noting high demand for tools that help compare products users may not know much about.
Key takeaways OpenAI has signed a multi-year, $10 billion agreement with AI chipmaker Cerebras Systems to secure computing infrastructure. The deal will deliver 750 megawatts of computing power throug...
KEY TAKEAWAYS: xAI implemented restrictions preventing Grok from editing images of real people in revealing clothing after global backlash California Attorney General Rob Bonta launched investigation...
Key takeaways TSMC reported fourth-quarter net income of NT$505.74 billion, beating analyst expectations and setting a new record. Fourth-quarter revenue reached NT$1.046 trillion, up 20.5% year over...
