Articles · How-to

AI-written fake reviews: how to spot them before they cost you money

By AI Week Daily editors · Published 14 November 2026 · 4 min read

Not financial advice. General consumer information only. Laws, platform policies and complaint channels change, so check the current rules where you live.

Star ratings decide what millions of people buy, and generative AI has made it cheap to write hundreds of believable reviews in minutes. Regulators have noticed. In the US a federal rule now explicitly bans fake reviews, including AI-generated ones, and India has a national standard for how platforms collect and publish reviews. Neither stops every fake from reaching your screen, so here is what the rules say, how to read reviews like a sceptic, and what to do if a fake review costs you money.

What the rules now say

United States. The Federal Trade Commission's Trade Regulation Rule on the Use of Consumer Reviews and Testimonials took effect on 21 October 2024 (Federal Register). It prohibits selling or buying fake reviews, buying positive or negative reviews, undisclosed reviews by company insiders, company-controlled "independent" review sites, certain ways of suppressing negative reviews, and selling or buying fake social-media influence such as bot followers. Coverage at the time noted the ban explicitly includes AI-generated reviews and reviews from people with no real experience of the product (TechCrunch). The maximum civil penalty was $51,744 per violation when the rule was issued, though courts can impose much less (Federal Register).

India. The Bureau of Indian Standards published IS 19000:2022, "Online Consumer Reviews — Principles and Requirements for their Collection, Moderation and Publication", which applies to any organisation publishing consumer reviews online, including sellers that collect reviews from their own customers (PIB, 12 December 2022). The Department of Consumer Affairs said compliance would initially be voluntary, and the standard describes ways to verify that a reviewer is real, such as email, phone or SMS checks and captchas, plus automated and manual moderation (PIB).

The practical takeaway: rules raise the cost of faking reviews and give you something to cite in a complaint, but they do not guarantee that every review you read is genuine.

Seven warning signs of a fake or AI-written review

  1. A burst of five-star reviews in a few days, especially soon after a listing goes live. Sort by newest and look at the dates.
  2. Generic praise with no specifics. "Amazing product, highly recommend, great quality" says nothing about size, fit, battery life or how it held up after a month. Real reviewers mention concrete details and small annoyances.
  3. The same phrases across different reviewers. AI-assisted fakes often reuse structure: an opening exclamation, three adjectives, a recommendation. Read five reviews in a row and look for the pattern.
  4. Reviews that describe a different product. Mentions of features the item does not have, or a different colour or model, can mean reviews were copied or merged from another listing.
  5. Reviewer profiles with no history or with dozens of five-star reviews for unrelated products on the same day.
  6. Ratings that do not match the text. Five stars attached to a lukewarm or off-topic comment is a red flag.
  7. Pressure to move off-platform. A review or seller message offering a refund, gift or "cashback" for a five-star rating is the kind of incentive the US rule bans when it is tied to a particular sentiment (TechCrunch).

A five-minute routine before any purchase over ₹2,000 or $50

  • Read the three-star reviews first. They are the least attractive to fake and usually the most specific about trade-offs.
  • Filter for verified purchases where the platform offers it, and check whether photos look like real homes rather than product shots.
  • Search the product name plus "problem" or "return" on a search engine or forum to find experiences outside the seller's control.
  • Compare the return window, not just the rating. A generous, clearly stated return policy protects you better than a 4.8-star average.
  • Screenshot the listing and reviews before you buy. If the product turns out to be very different from what reviews promised, you have evidence.

If a fake review cost you money

First, use the platform's own return or dispute process, and report the review using its "report" link, which helps moderation teams that standards like IS 19000 expect platforms to run. In India, if the seller or platform does not resolve it, register a complaint with the National Consumer Helpline, which takes e-commerce complaints by its toll-free number 1915, online, or through its app (Department of Consumer Affairs). In the US, you can report deceptive practices, including fake reviews, to the FTC at ReportFraud.ftc.gov.

Keep your order number, payment proof, screenshots of the reviews and the listing, and any messages from the seller. A clear paper trail is what turns "I was misled" into a refund.

Rule of thumb: trust detail, not stars. A handful of specific, mixed reviews from real-looking buyers is worth more than hundreds of identical five-star raves.

Sources

  1. Federal Register: Trade Regulation Rule on the Use of Consumer Reviews and Testimonials (22 August 2024)
  2. TechCrunch: FTC finalizes rule banning fake reviews, including those made with AI (14 August 2024)
  3. PIB: BIS issues standards for organizations publishing consumer reviews (12 December 2022)
  4. PIB: Department of Consumer Affairs framework on fake and deceptive reviews
  5. Department of Consumer Affairs: National Consumer Helpline
  6. FTC: ReportFraud.ftc.gov

Original article by AI Week Daily editors. Facts and figures are as stated by the linked sources on the publish date; check them for updates. Spotted an error? Tell us.

Not financial advice. General consumer information only. Laws, platform policies and complaint channels change, so check the current rules where you live.

More articles

All articles → · Daily digest →

Free daily digest

The AI news that matters, in five minutes.

5–10 hand-picked stories a day, each with a plain-language line on why it matters. Every item links to its source.

Free. One email a day, no spam. We only use your address to send AI Week Daily and never sell it. Unsubscribe any time with the link in every email, or reply to any email with “unsubscribe”. See the privacy note.

Free · Unsubscribe any time · Privacy

♥ Donate