OpenAI launches ChatGPT shopping agent that uses memory data to compare products
OpenAI just rolled out a new twist on ChatGPT: it now tries to act like a personal shopper, digging through the bits of data it’s stored about you, your browsing history, old searches, even hints that you were thinking about buying something. In theory it should take some of the grunt work out of comparing products and spotting price drops. In practice, though, it also hands a pretty detailed snapshot of your online habits over to the model.
Early users say the assistant can pull up past likes, flag a sale, or toss out a different brand that it thinks you might prefer. That sounds handy, but it also feels a bit like the line between helpful service and quiet surveillance is getting fuzzy. It’s unclear how far OpenAI will push this as a way to cash in, especially now that the company’s valuation is in the multibillion-dollar range.
The need to prove that price tag might nudge them toward more data-driven revenue ideas, making the trade-off between convenience and privacy harder to ignore.
"I know I'm not the product--I'm paying, and that's how the business model works."
I know I'm not the product--I'm paying, and that's how the business model works." With the new shopping agent using personal memory data to make purchase recommendations, that concern now feels more tangible. The pressure to justify OpenAI's massive valuation could push the company toward monetizing user data for targeted advertising or earning commissions on purchases. Deep Research, now optimized for shopping According to OpenAI, Shopping Research runs on a mini model post-trained on GPT-5-Thinking-mini.
The model was trained using reinforcement learning specifically for shopping tasks, with a focus on reading reliable sources and synthesizing information. It resembles OpenAI's existing deep research feature ("researches deeply across the internet") but tuned for e-commerce.
OpenAI’s new Shopping Research feature basically turns ChatGPT into a semi-autonomous product scout. It asks follow-up questions, pulls live data and then stitches together a visual guide. The tool leans on the model’s existing memory system, so suggestions feel a bit more personal than a plain search result.
The rollout is still in its infancy, and the interface looks more about convenience than flashy ads. Still, letting a bot remember your past queries raises obvious data-use concerns; the tagline “pay, not become the product” hints at a larger business-model debate. Critics point out that OpenAI’s lofty valuation might push the service toward deeper monetisation, but the exact plans are still fuzzy.
The visual guide could make comparison shopping easier, yet it’s hard to say whether the recommendations stay unbiased. As the feature grows, regulators and shoppers will probably start poking at how memory data shapes product picks. Until we see clear privacy safeguards and a transparent revenue strategy, confidence in the tool will likely rest on what we can actually observe.
Further Reading
- OpenAI Upgrades ChatGPT Search With Shopping Features - Slashdot
- OpenAI takes on Google, Amazon with new agentic shopping system - TechCrunch
- OpenAI launches shopping assistant - Semafor
- Introducing shopping research in ChatGPT - OpenAI
- OpenAI's new ChatGPT shopping tool promises 'in-depth' research - Modern Retail
Related Reading
Browse all OpenAI news →
LLMs & Generative AI OpenAI says AI saves knowledge workers 40‑80 minutes; use yields five‑fold gains
LLMs & Generative AI Grok Chat: AI for debugging, building, testing web apps with voice and images
LLMs & Generative AI Samsung adds Vision AI Companion, an AI Bixby, to TVs for real‑time queries
Related Topic Developers say Sora, unlike Vine/TikTok, is not about people in social media
Research & Benchmarks
Related Topic Jony Ive and Sam Altman unveil AI hardware prototype; OpenAI plans launch under two years
Business & Startups
Common Questions Answered
What does the new ChatGPT shopping agent do with a user’s personal memory data?
It sifts through browsing history, past queries, and purchase intent stored in ChatGPT’s memory to generate product comparisons and price alerts, helping users make buying decisions with less manual research.
How does the Shopping Research feature differ from traditional static search results?
Shopping Research runs on a mini model that asks follow‑up questions, pulls live data, and assembles a visual guide, leveraging the model’s memory to tailor suggestions rather than just returning a list of links.
What potential concerns are raised by OpenAI’s use of personal memory data in the shopping agent?
The integration places a trove of private browsing history and intent into the algorithm, sparking worries that OpenAI could monetize this data for targeted advertising or commissions, especially given its multibillion‑dollar valuation pressure.
When is the ChatGPT shopping agent expected to be widely available, and what is its current rollout status?
The feature is still in an early rollout phase, with the interface designed for convenience rather than overt advertising, and OpenAI has not yet announced a full public release date.