Observed risk
Warning signals found in the available evidence. This is not a measured probability that fraud has occurred.
Our methodology
Our evidence methodology and its limits
Fake Review Detector identifies supported warning signals and explains uncertainty. Its assessments are tools for further investigation, not declarations that a person, business, website or review is fraudulent.
Sources and inputs
The website checker uses bounded public website and domain observations, including available registration, DNS, TLS, page identity and policy evidence. Review analysis uses the review text, ratings, dates and reviewer details you supply. Other workflows accept the specific evidence described on their pages or import checks belonging to your workspace.
User-supplied observations, seller claims and community reports do not become independently verified facts merely because they are included in a report. A linked official source is useful only when it actually supports the relevant claim and its identity has been checked.
Risk, confidence and coverage
Confidence
How strongly the available evidence supports an interpretation. Sparse or conflicting inputs can weaken support.
Evidence coverage
How much relevant information is available for the assessment. Missing evidence stays unknown.
Review patterns and alternative explanations
Repeated wording, unusual timing, reviewer repetition and rating-text mismatches can justify a closer look. They may also have other explanations, including shared experiences, copied product terminology or an unrepresentative sample. A suspicious-pattern share is not a verified percentage of fake reviews.
Likewise, an old domain, HTTPS, a barcode, a verification badge or a company registration is only one piece of context. No single signal guarantees genuine products, reliable delivery or safe payment.
Dates, contradictions and next steps
Read the source and date attached to evidence when available. Websites, prices, identities and policies can change after a check. Preserve conflicting observations rather than treating more repeated copies of one claim as stronger independent support.
Before a purchase, independently verify the seller and payment recipient, confirm the final costs and terms, and check your provider's protection rules. If important evidence is missing, pause and obtain it rather than treating the absence of warnings as clearance.
Editorial approach and corrections
Guides explain practical checks and their limits. We avoid unsupported accuracy claims, fabricated fraud probabilities and claims of certification. User-authored or supplied material should be assessed in its own source context.
If a result or guide contains a specific error, submit the relevant evidence through support or the correction workspace. Explain which statement is affected and provide a source that addresses it.