AI & Media
How many fake reviews are posted every day?
A live count of fabricated product and service reviews posted to e-commerce and app platforms
Roughly 5.39 reviews every second.
Source: Amazon Transparency Report 2023; Fakespot/Mozilla; FTC rulemaking record. View on dashboard →
Why one in three Amazon reviews cannot be trusted
Fake reviews are ratings posted to mislead buyers. Businesses buy them to boost stars, or "review bomb" competitors. They come from review farms or AI. The FTC says they sway trillions in spending. Amazon removed 200+ million in 2021 alone. In 2024 the FTC finally banned them outright. Every platform fights it: Google Maps, Yelp, TripAdvisor, Trustpilot, app stores.
How fake reviews affect your everyday purchasing decisions
Most people read reviews before buying something online. Studies consistently find that 93% of consumers say online reviews influence their purchase decisions, and nearly 70% trust them as much as a personal recommendation. That trust has been systematically exploited.
Fakespot and Mozilla's analysis of Amazon product pages found that approximately 31% of reviewed products had ratings that were significantly inflated by fake reviews. That means when you filter for "4 stars and above," a substantial portion of what you see has been artificially elevated. You are paying real money based on fabricated social proof.
The industries worst affected are supplements, electronics accessories, beauty products, and budget clothing - exactly the categories where consumers most rely on reviews because they cannot inspect the product before buying. Knowing this, the practical advice is uncomfortable but accurate: if a product has 10,000 five-star reviews and no critical ones, be suspicious. Absence of negative feedback is itself a red flag.
The scale: nearly half a million fake reviews posted today alone
Amazon removed 170+ million fake reviews in 2023 (per Amazon's annual Transparency Report, as compiled by Fakespot), ~465,000 per day on one platform alone
FakespotIn 2024, the FTC issued its first formal rule explicitly banning fake consumer reviews and insider testimonials (FTC rulemaking record, cited by Fakespot)
FakespotFakespot analysis: 31% of Amazon reviews may be unreliable
FakespotFake reviews influence an estimated $152 billion in annual US consumer spending, per FTC 2024 rulemaking record compiled by Fakespot
FakespotAI-generated fake reviews grew dramatically after 2022, becoming increasingly hard to detect
FakespotPlatform crackdowns and regulatory responses: a timeline
- 2014Amazon bans incentivised reviews for the first time following FTC warnings
- 2016Amazon sues over 1,000 individuals for allegedly posting fake reviews for payment
- 2019FTC begins formal rulemaking; Yelp, Trustpilot, Google all report massive fake review removal campaigns
- 2024FTC issues formal final rule banning fake consumer reviews, first explicit regulatory prohibition
Fake review volume: from a niche problem to a systemic one
Fake review volumes grew steadily through the 2010s and accelerated sharply from 2022, as AI language models made it trivial to generate convincing product reviews at industrial scale with minimal effort.
| Year | Rate | Est. per day | Context |
|---|---|---|---|
| 2021 | 23K/hr | 548K | AI-assisted fake review generation begins |
| 2023-2024 | 19K/hr | 466K | FTC rule issued; AI generation widespread; removal data used as creation proxy |
| 2027 (forecast) | 29K/hr | 685K | AI-generated reviews commoditised; detection arms race |
Fake reviews vs. fake news: two parallel crises of fabricated content
Fabricated content shapes both what people buy and what people believe. Fake reviews distort purchasing decisions; fake news distorts political and social views. Both are accelerating with AI.
Fake reviews: how the $152 billion opinion economy is being manipulated
A multi-billion-dollar manipulation industry
Review farms in Southeast Asia and Eastern Europe sell realistic-looking reviews. Merchants swap fake reviews. Incentivised programmes blur the line. AI now generates them at scale. Trust in online reviews has cratered. 93% of buyers check reviews first, so the tactic works, and it's eating away at what trust remains.
The AI inflection point
Before 2022, generating convincing fake reviews at scale required low-paid human labour, which created natural cost constraints. Generative AI removed that constraint. Models can now produce thousands of contextually appropriate, grammatically varied, stylistically diverse fake reviews per hour at near-zero cost. Detection algorithms from platforms like Amazon and Yelp are locked in an escalating arms race with ever-more-sophisticated generation tools. The FTC's 2024 rule creates a legal deterrent, but enforcement against actors in non-US jurisdictions is limited, and AI lowers the cost of evasion as quickly as platforms raise the cost of detection.
Detection research: how platforms and academics fight fake reviews
| Year | Finding | Value | Source |
|---|---|---|---|
| 2019 | FTC begins formal investigation into fake review practices; initial rulemaking process starts | regulatory milestone | Fakespot |
| 2021 | Amazon: 200M+ fake reviews removed in 2021; launches brand transparency tools | 200M fake reviews removed (Amazon, 2021) | Fakespot |
| 2023 | Amazon: 170M+ fake reviews removed; Fakespot finds 31% of reviews unreliable on the platform | 170M fake reviews removed (Amazon, 2023) | Fakespot |
| 2024 | FTC issues formal rule banning fake reviews; AI-generated reviews increasingly prevalent | regulatory milestone | Fakespot |
In perspective
A fake review is created roughly every five seconds. Most shoppers take longer than that to read a real one.
If someone read through every fake review Amazon removed in 2023 at one per minute, it would take over 320 years.
Amazon removed 170 million fake reviews in 2023. That's more people than live in Germany and France combined.
How the number is calculated
Amazon removed 170 million+ fake reviews in 2023 (Amazon Transparency Report) – a confirmed lower bound on how many were posted. 170,000,000 ÷ 365 ≈ 465,000 detected/removed per day from Amazon alone. Adding other platforms (Google Maps, Yelp, Trustpilot, App Store) and estimating undetected fakes, the global daily total of fake reviews posted is estimated at ~19,400/hour. The live counter uses 19,400/hour ÷ 3,600 ≈ 5.4/sec, using detected removals as a proxy for posts.
Sources: Fakespot - Fake Review Analysis. Methodology →
Frequently asked questions
- How many fake reviews are posted per day?
- Estimates vary widely. Amazon reported removing 170+ million fake reviews in 2023 alone, implying approximately 465,000 per day on one platform. Including other platforms (Google, Yelp, Trustpilot, etc.), the global daily total is estimated in the millions.
- What percentage of reviews are fake?
- Studies and platform reports suggest 15-30% of reviews on major e-commerce and service platforms may be inauthentic. A 2023 Fakespot analysis found that 31% of Amazon reviews were unreliable. ReviewMeta estimated similar rates.
- What is the economic impact of fake reviews?
- The FTC estimates fake reviews influence over $152 billion in annual US consumer spending. Globally, the figure is much higher. Businesses pay millions for fake review campaigns; competitors suffer unfair disadvantage.
Why trust this data
The Amazon removal figure comes from Amazon's annual Transparency Report, published each April. Fakespot's independent analysis (acquired by Mozilla in 2023) provides the 31% unreliability rate for Amazon reviews. FTC data on the economic impact comes from the agency's 2024 rulemaking record, which drew on peer-reviewed economics research.
Sources
Explore related: Fake news shared - AI-generated news - Online scam victims, and the live AnythingCounter dashboard.