The policy covers user reviews of apps in the Meta Horizon Store. Meta says it welcomes user feedback so users can understand the app experience and developers can hear from users directly. The document gives its date of last revision as July 29, 2024.
Supporting quotations
User reviews are available for Meta Horizon Store apps.
We welcome user feedback on apps so that users can understand the app experience and developers can hear from users directly.
Date of last revision: July 29, 2024
Qualifications & exceptions
Covers reviews of Meta Horizon Store apps; the page does not say whether it also governs reviews anywhere else.
The page states a date of last revision (July 29, 2024), not an effective date.
Not stated here
The document does not say when the policy took effect, or whether a later revision exists.
Meta reserves the right to take action on a user review, up to and including removing it, if the review falls into one of the listed categories: abusive or inflammatory content, irrelevant reviews, fake or manipulative reviews, incentivized or financially interested reviews, or reviews that disclose personal information. The sentence says action goes 'up to and including' removal, so removal is the stated upper limit; the document names no other action.
Supporting quotations
Meta reserves the right to take action on user reviews, up to and including removal of the review, if the review is:
Qualifications & exceptions
Only where a review meets one of the categories listed in the policy.
'Reserves the right' is permission for Meta, not a promise that Meta will act or review every post.
Not stated here
The document does not say what actions short of removal Meta may take.
It does not say whether Meta notifies the reviewer, how Meta decides, or whether a reviewer can appeal.
Inflammatory, hateful or abusive reviews, and Code of Conduct violations
A review may be acted on if it is inflammatory, hateful, contains abusive language that targets a particular group of people, or violates Meta's Code of Conduct for Virtual Experiences. The Code of Conduct is a linked page that this document does not reproduce, so what it prohibits is not stated here.
The phrase 'that targets a particular group of people' follows 'abusive language'; the wording does not make clear whether it also limits 'inflammatory' or 'hateful', and the document does not resolve that.
Not stated here
The Code of Conduct for Virtual Experiences is only linked, not supplied, so its rules are not part of this analysis.
The document does not define 'inflammatory' or 'hateful'.
Content & IP Rights · Restriction or exception
Reviews not based on direct experience or about the platform
A review may be acted on if it is irrelevant. The policy says this includes a review not based on the user's direct experience with the app, a review unrelated to the intended app experience, and reviews focused on the platform rather than the app experience.
Supporting quotations
irrelevant, including where the review is not based on the user’s direct experience with the app or is unrelated to the intended app experience, including reviews focused on the platform rather than the app experience.
Qualifications & exceptions
The three examples are introduced by 'including', so the list is not stated to be exhaustive.
Not stated here
The document does not say how Meta decides whether a reviewer had direct experience with the app.
Content & IP Rights · Restriction or exception
Fake, spam or manipulative reviews, including review bombing
A review may be acted on if it is fake, spam or engagement bait, or is posted to artificially manipulate an app's rating or as part of a coordinated attack on an app. The policy gives review bombing and brigading as examples.
Supporting quotations
fake, spam, engagement bait, or posted to artificially manipulate the app’s rating or as part of a coordinated attack on an app (e.g. review bombing, brigading, etc).
Qualifications & exceptions
Review bombing and brigading are given as examples ('e.g.'), not as a complete list.
Not stated here
The document does not define 'fake' or 'engagement bait' or say how Meta detects coordinated activity.
Content & IP Rights · Restriction or exception
Incentivized reviews and reviewers with a financial interest
A review may be acted on if it is incentivized, for example a review left to unlock a rare object in a game, or if the reviewer has a financial interest, for example paid reviews or reviews containing referrals.
Supporting quotations
incentivized (e.g. leave a review to unlock a rare object in the game) or the reviewer has a financial interest (e.g. paid reviews or reviews containing referrals).
Qualifications & exceptions
The examples (unlocking a rare object, paid reviews, referrals) are introduced by 'e.g.' and are not stated to be the only cases.
Not stated here
The document does not say how Meta learns that a review was incentivized or paid.
Content & IP Rights · Restriction or exception
Reviews that disclose personal information
A review may be acted on if it discloses personal information such as age, passwords, an email address, a phone number or the like.
Supporting quotations
disclosing personal information such as age, passwords, email address, phone number or the like.
Qualifications & exceptions
The listed items are examples ('such as ... or the like'), not a complete list.
Not stated here
The document does not say whether the information must be the reviewer's own or someone else's.
Use & Processing · Company practice
Meta AI may summarize review text, and summaries may be wrong
Meta may process the text of one or more user reviews for an app with Meta AI to generate a summary for the app. Meta warns that such summaries may not be representative of all user reviews, and that some may be inaccurate or inappropriate. A link points to a Meta privacy page about generative AI, which this document does not reproduce.
Supporting quotations
Text from one or more user reviews for an app may be processed by Meta AI to generate a summary for the app.
Such summaries may not be representative of all user reviews, and some may be inaccurate or inappropriate.
Qualifications & exceptions
'May be processed' states a possibility, not that every review is processed.
Covers text from one or more user reviews for an app, used to generate a summary for that app.
Not stated here
The linked Meta AI privacy page is not supplied, so any detail on how review text is used, kept or shared is not part of this analysis.
The document does not say whether reviewers can opt out of their review text being summarized.