There was a time when online shopping meant typing two or three words into a search box and scrolling through pages of results. "Running shoes." "Air fryer." "Phone under $300." You did the comparing, the reading, the worrying about whether the price was fair.
That is changing fast. A new kind of helper has moved into the world's biggest stores: the AI shopping assistant. Instead of keywords, you talk to it the way you would talk to a knowledgeable shop assistant. You ask questions, it answers, compares, and sometimes even watches prices for you.
Amazon, Google and OpenAI have all launched their own versions. Hundreds of millions of people already use them. If you shop online, it is worth understanding what these assistants can do, where they shine, and where you should not trust them.
Meet your new shopping assistant
An AI shopping assistant is a chatbot built for buying. You describe what you need in plain language — "I need durable running shoes for wide feet, under $100" — and it searches the store's catalog, reads customer reviews, and gives you a shortlist with explanations.
The difference from normal search is the conversation. You can ask follow-up questions: "Which of these is best for flat feet?" "Is the waterproof version worth the extra $20?" The assistant remembers the context and narrows things down, instead of making you start over with new keywords.
These assistants are trained on product catalogs, customer reviews, questions other shoppers asked, and information from across the web. They are genuinely useful for the most boring part of shopping: the research.
Amazon's assistant: from Rufus to Alexa for Shopping
Amazon was first among the giants. It launched its AI shopping assistant, called Rufus, as a test inside its shopping app in February 2024. Shoppers could ask it anything from broad questions ("what should I consider when buying a coffee maker?") to specific ones ("is this jacket machine washable?").
Rufus grew quickly. Amazon says more than 300 million customers used it during 2025, and the company credits its AI shopping assistant with nearly $12 billion in extra sales that year. Over 50 upgrades turned it from a simple question-answer tool into something closer to a personal shopping agent.
Then came a name change many shoppers missed. In May 2026, Amazon folded Rufus together with its Alexa+ assistant and renamed the result Alexa for Shopping in the United States. In the UK, it is still called Rufus. Same job, new name — and an icon for it now sits on every product page.
What can it do today? Quite a lot:
- Price history. Ask "is this the lowest price recently?" and it shows how the price moved over the last 30 or 90 days. A full one-year view is rolling out in the US, the UK and India.
- Price alerts. Tell it the price you would pay, and it pings you when the item drops there.
- Comparisons. It lines products up side by side and summarizes what reviewers say are the pros and cons.
- Auto Buy (Prime members). Set a target price and it buys the item automatically when the price hits it. You can cancel within 24 hours.
- Buy for Me. In the US, the assistant can even complete purchases on other retailers' websites for you.
The price history feature alone is worth knowing about. "Deal" labels are not always honest — a product "reduced" from an inflated price is no bargain. A year of price history exposes that trick in seconds.
ChatGPT joins the mall
In April 2025, OpenAI added shopping features to ChatGPT. Ask it for "the best noise-cancelling headphones under $200" and it returns visual product cards with images, prices, star ratings and short summaries of what reviewers say, plus direct links to buy.
Two things make ChatGPT's version interesting. First, it works for everyone, including free users and people who are not even logged in. Second, OpenAI says the picks are not ads and it earns no commission on sales — the recommendations come from structured product data like prices, descriptions and reviews. It started with fashion, beauty, home goods and electronics.
The experience is conversational rather than keyword-driven. Instead of scrolling through pages of results, you describe your situation and refine it: "something lighter," "cheaper," "better for small kitchens." For shoppers who find traditional search overwhelming, that is a real improvement.
Google answers before you click
Google took a different path. Instead of building a separate shopping chatbot, it put AI answers directly on top of search results. AI Overviews — short AI-written summaries — launched in the US in May 2024 and expanded to more than 200 countries and 40 languages by May 2025. Google says over 2 billion people see them every month.
Search for a product and you may now get an AI summary of the best options before you ever click a link. In 2026, Google merged this with its AI Mode, so you can ask follow-up questions and keep the conversation going right inside search.
This matters for shopping because product research increasingly happens without visiting a single store page. Research by Pew found that when an AI summary appears, users click a traditional search result in only 8% of visits, compared with 15% when there is no summary. The answer itself has become the destination.
How this changes the way we buy
Step back and the pattern is clear. Shopping is moving from searching to asking. Three shifts stand out:
1. Less tab-hopping. The classic routine — open ten tabs, compare specs, read reviews on three sites — gets compressed into one conversation. The assistant does the reading and hands you the summary.
2. Price transparency. Price history and alerts shift power toward the buyer. Stores can no longer quietly raise a price the week before a "sale." If shoppers check history before buying, fake discounts stop working.
3. The agent era. The newest assistants do not just advise — they act. Watching a price, reordering essentials, even buying at your target price. Shopping becomes something you supervise rather than something you do click by click.
For busy people, that is genuinely convenient. For careful spenders, the price tools are the killer feature. But convenience has a price, which brings us to the risks.
7 smart ways to use AI shopping assistants
You will get far more out of these tools with a few habits:
- Be specific about your situation. "Headphones for a noisy office, under $150, comfortable for glasses wearers" beats "good headphones." The more context, the better the shortlist.
- Ask for comparisons, not winners. "Compare these three on battery life and comfort" gives you a useful table. "Which is best?" gives you an opinion you cannot check.
- Always verify the price on the retailer's page. AI answers can lag behind real prices, especially during sales. Click through before you pay.
- Check the price history before any "deal." On Amazon, ask the assistant for the price history. A 40%-off badge means nothing if the price was lower last month.
- Read a few real reviews yourself. AI summaries are great for the overall picture, but the one-star reviews often reveal the specific flaw that would bother you.
- Watch for sponsored content. Amazon now shows Sponsored Products inside its assistant. An AI answer is not automatically neutral — check whether a recommendation is paid placement.
- Keep auto-buy on a short leash. Price alerts are safe; automatic purchases are not. If you use Auto Buy, set conservative target prices and check your orders regularly.
What online sellers should know
If you sell online, these assistants are your new gatekeepers. When an AI answers a shopper's question, it picks which products to mention — and most shoppers will never scroll past its answer.
The sellers who win share a pattern: clear product titles, detailed and honest descriptions, plenty of genuine customer reviews, and complete product information (sizes, materials, compatibility). AI assistants quote from reviews and product pages, so thin listings with two-sentence descriptions simply do not get recommended.
Answering common questions in your listing helps too. If shoppers keep asking "is it waterproof?" and your description answers it, the assistant can quote you directly. Think of your product page as something written for both humans and AI readers now.
The risks nobody should ignore
AI shopping assistants are impressive, but they are not oracles. A few honest warnings:
They can be wrong. Amazon's own help page warns that its assistant "may not always get things right." Prices change, stock runs out, and AI sometimes mixes up product versions. Treat every answer as a starting point, not the final word.
They can nudge you to spend. An assistant that remembers your tastes and pings you about price drops is also an assistant that creates buying urges you did not have. Convenience shopping can quietly become impulse shopping.
They know a lot about you. Personalized recommendations need personal data — your searches, purchases, and preferences. Check the privacy settings and decide what you are comfortable sharing.
Auto-buy needs supervision. Letting software spend your money is a big step. Start with alerts, not automatic purchases, and set spending limits wherever the app allows.
Frequently asked questions
Are ChatGPT's product recommendations ads?
No — at least not currently. OpenAI says ChatGPT's shopping results are chosen independently and it earns no commission on sales. That said, the company has discussed advertising as a future possibility, so it is worth staying alert to changes.
What happened to Amazon Rufus? Is it gone?
Not gone — renamed. In May 2026, Amazon renamed Rufus to Alexa for Shopping in the United States and combined it with Alexa+ features. In the UK it is still called Rufus. Your history and features carried over.
Do I need Prime to use Amazon's AI shopping assistant?
No. The assistant itself is free for anyone signed in to Amazon. Prime is only needed for certain features like Auto Buy and for shopping Prime-exclusive sale events.
Can these assistants really buy things for me?
Yes, within limits. Amazon's Auto Buy (Prime) purchases an item when it hits your target price, and its Buy for Me feature can check out on other retailers' sites in the US. ChatGPT and Google currently link you to the store to complete the purchase yourself.
Should I trust the prices an AI assistant shows me?
Mostly, but verify. AI answers can be out of date, especially during fast-moving sales. Always click through to the retailer's page to confirm the current price before paying — it takes ten seconds and has saved many shoppers from overpaying.
The way we shop is being rewritten by conversation. The shoppers who thrive will be the ones who use these assistants as brilliant research helpers — while keeping the final decision firmly in their own hands.
