Reddit Tests AI Product Reviews Built From Community Comments

Reddit announced on November 10, 2026, that it is testing Reddit Reviews, an AI-organized shopping feature that consolidates eligible community discussions into product pages with sentiment summaries, recurring themes and original-comment links. The initial test covers about 10,000 frequently discussed products.

The feature is rolling out over the next few weeks to US users on iOS, Android and the web, starting with Electronics, Health & Beauty, Home & Garden and Fashion & Apparel. Reddit plans to add products and categories and expand beyond the US during 2027.

Review Pages Keep Original Discussions in View

Users can open a review through a product or category search, or by selecting a highlighted product name on a post’s detail page. A specific product can produce a review card at the top of search results, while broader queries can return a row of cards.

A What people are saying snapshot divides the included opinions into positive, mixed or neutral and negative. Even small groups remain visible, so the colored sections do not represent exact percentages. Recurring themes identify strengths, drawbacks and practical considerations such as comfort or ease of use.

A filterable Mentions feed exposes the comments behind the review, and selecting one opens the original conversation. Pages can also highlight commonly compared products and link to category views organized around needs expressed in discussions, including value, budget and ease of use.

When no dedicated review is available, Reddit’s AI Search can answer the product question from Reddit conversations. The AI Search documentation says those responses include inline citations to original posts and comments.

AI Eligibility Rules and Accuracy Limits

The Reddit Reviews documentation says large language models identify and organize comments that appear to describe personal product experiences. The models classify qualifying conversations as positive, negative or mixed, consider surrounding discussions for context, and assess tone, product attributes, helpfulness and suitability for display. They do not generate or rewrite the user comments shown.

The initial reviews use only English-language conversations from US users. Eligible comments must also come from the past year, be safe for work and sit outside mature communities. Verified purchases and evidence of long-term ownership are not required; first impressions and in-store trials can qualify. Questions alone, speculation, secondhand claims, sale listings and mentions without an experience-based opinion do not.

Summaries may draw on only a selection of conversations, and Reddit cautions that AI can miss context or make errors. It does not independently verify that commenters used or bought the products; nor does it test, certify or endorse the products themselves. Users also cannot currently hide Reddit Reviews or opt eligible comments out of inclusion.

Earlier Shopping Tests and Merchant Controls

Reddit Reviews extends a series of shopping tests that began with a related-products carousel in AI search for a small group of US users searching for electronics in February 2026. An expanded experience rolling out from June 3, 2026 added AI-powered product deep dives with direct quotes and product tabs on business profiles. Testing reached Home & Garden and Fashion & Apparel on September 10, 2026.

Brand and retailer catalogs can add images, prices and purchase links, while merchants cannot pay for product inclusion or removal of negative opinions. The catalog material does not change the user opinions shown, and prices may be outdated if merchants have not updated their information. Purchases are completed on merchant websites; Reddit does not process the transactions and receives no commissions.

Reddit says volunteer moderators enforcing community rules, its Safety team enforcing sitewide rules and user voting together help limit spam and inauthentic activity in the conversations feeding reviews.

That gives Reddit a structured discovery product built from existing communities while keeping paid placement out of the sentiment itself. Its central constraint as coverage expands is evidentiary: the pages summarize selected, unverified experiences, making the quality of comment selection and access to full discussions essential to their usefulness.