Facebook and Instagram may recommend content, accounts and entities you do not already follow
The document says "we" make personalized recommendations to help people discover new communities and content. Both Facebook and Instagram may recommend content, accounts, and entities (such as Pages, Groups, or Events) that people do not already follow. It names Pages You May Like, "Suggested For You" posts in Feed, People You May Know, and Groups You Should Join as examples of these recommendation experiences.
Supporting quotations
We make personalized recommendations to the people who use our services to help them discover new communities and content.
Both Facebook and Instagram may recommend content, accounts, and entities (such as Pages, Groups, or Events) that people do not already follow.
Some examples of our recommendation experiences include Pages You May Like, “Suggested For You” posts in Feed, People You May Know, or Groups You Should Join.
Qualifications & exceptions
The recommended items are ones the person does not already follow.
The named experiences are examples ("include"), not a complete list.
Not stated here
The page does not say where each recommendation experience appears or list every one.
The document does not name the company behind "we"; it speaks as the provider of Facebook and Instagram.
Recommendations are personalized for each person, for example by Friends in common
The stated goal is recommendations that are relevant and valuable to each person who sees them. The document says it works toward that goal by personalizing recommendations, "making unique recommendations for each person". Its one example is that if you and another person have Facebook Friends in common, it may suggest that person as a potential new Friend for you.
Supporting quotations
Our goal is to make recommendations that are relevant and valuable to each person who sees them.
We work towards our goal by personalizing recommendations, which means making unique recommendations for each person.
if you and another person have Facebook Friends in common, we may suggest that person as a potential new Friend for you.
Qualifications & exceptions
The Friends-in-common example is worded "may suggest": a suggestion is possible, not promised.
Not stated here
Friends in common is the only signal the page gives; it does not list other information used to personalize recommendations.
Content & IP Rights · Company practice
Recommendations Guidelines set a higher standard than the Community Standards
Facebook says it has guidelines that govern what content it will recommend, and that through these Recommendations Guidelines it works to avoid recommending content that could be low-quality, objectionable, or particularly sensitive. The guidelines are designed to maintain a higher standard than the Community Standards because recommended content and connections come from accounts or entities the person has not chosen to follow. As a result, not all content allowed on the platform is eligible for recommendation. Separately, the page says more severe Community Standards violations will be removed from Facebook.
Supporting quotations
At Facebook, we have guidelines that govern what content we will recommend to people.
we work to avoid making recommendations that could be low-quality, objectionable, or particularly sensitive.
Our Recommendations Guidelines are designed to maintain a higher standard than our Community Standards, because recommended content and connections are from accounts or entities you haven't chosen to follow. Therefore, not all content allowed on our platform will be eligible for recommendation.
As always, more severe violations of our
will be removed from Facebook.
Qualifications & exceptions
The higher standard is explained by recommended content and connections coming from accounts or entities the person has not chosen to follow.
Being allowed on the platform does not make content eligible for recommendation.
Removal is stated for "more severe" Community Standards violations, not for everything that fails these Recommendations Guidelines.
Not stated here
The Community Standards are linked but not supplied here, and the page does not say what makes a violation "more severe".
Facebook uses technology to detect and avoid recommending items that do not meet the guidelines
Facebook says it uses technology to detect content, accounts, and entities that do not meet the Recommendations Guidelines and to help it avoid recommending them.
Supporting quotations
We use technology to detect content, accounts, and entities that don’t meet these Recommendations Guidelines and to help us avoid recommending them.
Qualifications & exceptions
The technology is described as helping to "avoid" recommending these items, not as a guarantee that none is ever recommended.
Not stated here
The page does not describe the technology, say whether people also review items, or say how accurate detection is.
Under-18 users get additional age-appropriate guidelines and enhanced protections
For users under 18, Facebook says it has additional age-appropriate content guidelines. It says it worked with youth experts around the world to understand what types of content may be appropriate for adults but too mature for teens, and that its guidelines are designed to ensure recommendations are appropriate for its global community while providing enhanced protections for teen users.
Supporting quotations
For users under 18, Facebook has additional
We’ve worked with youth experts around the world to understand what types of content may be appropriate for adults, but too mature for teens.
Our guidelines are designed to ensure recommendations are appropriate for our global community while providing enhanced protections for teen users.
Qualifications & exceptions
The additional guidelines apply to users under 18.
"Designed to ensure" describes the aim of the guidelines, not a measured result.
Not stated here
The age-appropriate content guidelines are linked but not supplied here.
The page does not say how a user's age is determined.
The guidelines were developed with expert input and are published to inform people and content creators
Facebook says it sought input from 50 leading experts specializing in recommender systems, expression, safety, and digital rights while developing the guidelines. It calls those consultations part of its constant efforts to improve the guidelines and to give people a safe and positive experience when they receive recommendations. It says it provides this information about the content, accounts and entities it tries to avoid recommending to keep its community informed and to give guidance to content creators.
Supporting quotations
In developing these guidelines, we sought input from 50 leading experts specializing in recommender systems, expression, safety, and digital rights.
Those consultations are part of our constant efforts to improve these guidelines and provide people with a safe and positive experience when they receive recommendations on our platform.
We want to provide you with more information about the types of content, accounts, and entities that we try to avoid recommending, both to keep our community more informed and to provide guidance for content creators about recommendations.
Qualifications & exceptions
The expert input is described as part of ongoing improvement of the guidelines.
Not stated here
The experts, the input they gave and how it changed the guidelines are not described.
Three categories of allowed content may not be eligible for recommendations
The document says there are three categories of content that are allowed on its platforms but may not be eligible for recommendations, and gives illustrative examples of each. The categories are content that impedes its ability to foster a safe community, content associated with low-quality publishing, and false or misleading content.
Supporting quotations
There are three categories of content that are allowed on our platforms, but that may not be eligible for recommendations.
Content that impedes our ability to foster a safe community, such as:
Content that is associated with low-quality publishing, such as:
False or Misleading Content, such as:
Qualifications & exceptions
"May not be eligible" means eligibility is not ruled out for every item in these categories.
The examples in each category are described as illustrative.
Not stated here
The page does not say which allowed items in these categories are still recommended.
Self-harm, suicide, eating disorder, death and depression content may not be recommended; some is hidden from teens
Under the safe-community category, the document lists content that discusses self-harm, suicide, or eating disorders, and content that depicts or trivialises themes around death or depression. It says it removes content that encourages suicide or self-injury, or any graphic imagery. It also says it hides certain suicide and self-injury related content to protect teens from potentially distressing or sensitive material, and gives age-restriction examples: a post describing someone's own experiences with suicide, self-injury or eating disorders (except in the context of recovery), a photo or video of people in a hospital engaging in euthanasia or assisted suicide, and a Story about someone's own extreme weight loss behavior.
Supporting quotations
Content that discusses self-harm, suicide, or eating disorders, as well as content that depicts or trivialises themes around death or depression.
We hide certain suicide and self-injury related content to protect teens from potentially distressing or sensitive material.
A post where someone is describing their own personal experiences with suicide, self-injury or eating disorders, except in the context of recovery.
A photo or video showing people in a hospital engaging in euthanasia or assisted suicide.
A Story where someone is talking about their own extreme weight loss behavior.
Qualifications & exceptions
Removal is stated for content that encourages suicide or self-injury, or any graphic imagery, not for all discussion of these topics.
The teen hiding applies to "certain" content, and the first example excludes posts about recovery.
The examples are introduced as what the publisher "would age restrict".
Not stated here
"Age restrict" is not defined; the page does not say whether it means hidden from teens entirely or kept out of teens' recommendations.
The page does not say how this content is treated for adults beyond the category statement that it may not be eligible for recommendation.
Content & IP Rights · Restriction or exception
Content that may depict violence may not be recommended; most graphic imagery is hidden from teens
Under the safe-community category, the document lists content that may depict violence, such as people fighting. It says it removes graphically violent content. It says it hides most graphic and disturbing imagery from teens, even if it would allow it behind a warning screen for adults. Its age-restriction examples are a photo of a severely burned person (which it would cover with a warning screen for adults) and a photo or video of shootings, brutality, or deadly car crashes.
Supporting quotations
Content that may depict violence, such as people fighting.
We hide most graphic and disturbing imagery from teens, even if we’d allow it behind a warning screen for adults.
A photo of a severely burned person, which we’d cover with a warning screen for adults.
A photo or video of shootings, brutality, or deadly car crashes.
Qualifications & exceptions
Removal is stated for graphically violent content.
For adults, some graphic content may be allowed behind a warning screen; for teens, most is hidden.
Not stated here
"Age restrict" is not defined on the page.
The page does not say which graphic content adults see behind a warning screen and which is removed.
Content & IP Rights · Restriction or exception
Sexually explicit or suggestive content may not be recommended; sexually suggestive content is hidden from teens
Under the safe-community category, the document lists content that may be sexually explicit or suggestive, such as pictures of people in see-through clothing. It says it removes content that contains adult nudity or sexual activity. It says it hides additional content from teens that does not contain explicit nudity or sexual activity but could be considered sexually suggestive. Its age-restriction examples are a post with near nudity such as nudity covered by a digital overlay, a photo zoomed in on someone's buttocks, and a Story with sexually suggestive language describing sexual encounters, including sexual metaphors.
Supporting quotations
Content that may be sexually explicit or suggestive, such as pictures of people in see-through clothing.
We hide additional content from teens that doesn’t contain explicit nudity or sexual activity but could be considered sexually suggestive.
A post with near nudity such as nudity covered by digital overlay.
A photo zoomed in on someone’s buttocks.
A Story with sexually suggestive language describing sexual encounters, including the use of sexual metaphors.
Qualifications & exceptions
Removal is stated for content that contains adult nudity or sexual activity.
The teen hiding is for content without explicit nudity or sexual activity that "could be considered" sexually suggestive.
Not stated here
"Age restrict" is not defined on the page.
The page does not say who decides what could be considered sexually suggestive.
Content & IP Rights · Restriction or exception
Content promoting regulated products may not be recommended; some is hidden from teens
Under the safe-community category, the document lists content that promotes the use of certain regulated products, such as tobacco or vaping products, adult products and services, or pharmaceutical drugs. It says it removes content that attempts to sell or trade most regulated goods. It says it hides content from teens that could influence them to engage in activities that are potentially harmful. Its age-restriction examples are a post offering to sell tobacco, nicotine products, alcohol, or firearms when shared by a legitimate brick-and-mortar business, a Story encouraging people to take psychedelic drugs or cannabis products, images with recipes for alcoholic drinks, and a post selling weight loss products or services. A further item covers content that promotes or favorably depicts the use of weight loss products or services and cosmetic surgeries.
Supporting quotations
Content that promotes the use of certain regulated products, such as tobacco or vaping products, adult products and services, or pharmaceutical drugs.
We hide content from teens that could influence them to engage in activities that are potentially harmful.
A post offering to sell tobacco, nicotine products, alcohol, or firearms when shared by a legitimate brick-and-mortar business.
A Story encouraging people to take psychedelic drugs (also known as entheogens) or cannabis products.
Images with recipes for alcoholic drinks.
A post selling weight loss products or services.
Content that promotes or favorably depicts the use of weight loss products or services and cosmetic surgeries.
Qualifications & exceptions
Removal is stated for content that attempts to sell or trade "most" regulated goods, so some may remain.
The teen hiding is for content that "could influence" teens toward potentially harmful activities.
The first example concerns posts shared by a legitimate brick-and-mortar business.
Not stated here
The extracted list does not show whether the weight-loss and cosmetic-surgery item is a further age-restriction example or a separate listed item; the card states it as listed.
"Age restrict" is not defined on the page.
The page does not say which regulated goods are not covered by removal.
Content & IP Rights · Restriction or exception
Content shared by a non-recommendable account or entity may not be recommended
Content shared by any non-recommendable account or entity, for example Groups or Pages, is listed in the safe-community category of content that may not be eligible for recommendation. The document says these accounts and entities are outlined below, in the account and entity recommendations section.
Supporting quotations
Content shared by any non-recommendable account or entity (e.g., Groups or Pages, as outlined below).
Qualifications & exceptions
This applies to content shared by an account or entity that is itself non-recommendable.
Groups and Pages are given as examples ("e.g."), not a full list.
Not stated here
Which accounts and entities are non-recommendable is set out in the section below, not in this item.
Unoriginal content and engagement-manipulation tactics may not be recommended
The low-quality publishing category lists unoriginal content that is largely repurposed from another source without adding material value, and long captions unrelated to the underlying content together with coordinated comment networks intended to artificially drive engagement and distribution.
Supporting quotations
Content that is associated with low-quality publishing, such as:
Unoriginal content that is largely repurposed from another source
Long captions unrelated to the underlying content and coordinated comment networks intended to artificially drive engagement and distribution.
Qualifications & exceptions
Repurposed content is covered when it is "largely" repurposed and adds no material value.
The comment-network item is limited to networks "intended to" artificially drive engagement and distribution.
Not stated here
The page that explains "without adding material value" is linked but not supplied here.
Content & IP Rights · Restriction or exception
In countries with third-party fact-checking, content found false by independent fact-checkers may not be recommended
The false or misleading content category lists, in countries where Meta supports third-party fact-checking, content including claims that have been found false by independent fact-checkers. The document says it removes misinformation that could cause physical harm or suppress voting.
Supporting quotations
False or Misleading Content, such as:
In countries where Meta supports third-party fact-checking, content including claims that have been found false by independent fact-checkers.
We remove misinformation that could cause physical harm or suppress voting.
Qualifications & exceptions
The category is tied to countries where Meta supports third-party fact-checking.
Removal is stated only for misinformation that could cause physical harm or suppress voting.
Not stated here
The page does not list the countries where Meta supports third-party fact-checking.
The page does not say what applies to false claims in other countries.
Accounts and entities with recent violations or repeated non-recommendable content are generally not recommended
The document says it generally does not recommend accounts (including Profiles and Page admins) or entities (such as Pages, Groups, or Events) that recently violated its Community Standards, or that repeatedly and/or recently shared content it tries not to recommend across the categories in the Content Recommendations section, including the names or cover photos associated with Groups or Pages. The first criterion does not include accounts or entities that the publisher otherwise removes from its platforms for violating the Community Standards.
Supporting quotations
We generally do not recommend accounts (including Profiles and Page admins) or entities (such as Pages, Groups, or Events) that:
Recently violated our Community Standards.
This does not include accounts or entities that we otherwise remove from our platforms for violating our Community Standards.
Repeatedly and/or recently shared content (including the names or cover photos associated with groups or Pages) we try not to recommend across the categories described in the Content Recommendations section above.
Content recommendations
Qualifications & exceptions
The rule is stated as "generally", so exceptions are possible.
The first criterion turns on a "recent" violation; the second on content shared "repeatedly and/or recently".
Accounts or entities removed for violating the Community Standards are outside the first criterion.
Not stated here
"Recently" and "repeatedly" are not defined, so the page does not say how many items or how long a period triggers the rule.
The page does not say how long an account or entity stays non-recommendable.
Accounts and entities that repeatedly use misleading practices to build followings are generally not recommended
Among the accounts and entities the document says it generally does not recommend are those that repeatedly engaged in misleading practices to build followings, such as purchasing ‘likes’.
Supporting quotations
We generally do not recommend accounts (including Profiles and Page admins) or entities (such as Pages, Groups, or Events) that:
Repeatedly engaged in misleading practices to build followings, such as purchasing ‘likes’.
Qualifications & exceptions
The rule is stated as "generally".
It applies to repeated misleading practices; purchasing likes is given as an example ("such as").
Not stated here
The page does not list other misleading practices or say how many occurrences count as "repeatedly".
Accounts and entities that repeatedly posted fact-checked false or altered information are generally not recommended
Where Meta supports third-party fact-checking, accounts and entities that recently and repeatedly posted "False" and "Altered" information, as determined by independent third-party fact-checkers, are among those the document says it generally does not recommend.
Supporting quotations
We generally do not recommend accounts (including Profiles and Page admins) or entities (such as Pages, Groups, or Events) that:
In countries where Meta supports third-party fact-checking, recently and repeatedly posted “False” and “Altered” information as determined by independent third-party fact-checkers.
Qualifications & exceptions
Applies in countries where Meta supports third-party fact-checking.
Both "recently" and "repeatedly" are required in this item.
The rule is stated as "generally".
Not stated here
The page does not list the countries where Meta supports third-party fact-checking, define "recently", or say how many posts are "repeatedly".
Teens are not recommended accounts that regularly share age-inappropriate content or look inappropriate for teens
The document says teens will not be recommended accounts that it has found regularly share age-inappropriate content, or accounts whose name or bio suggests the account is inappropriate for teens.
Supporting quotations
Teens will not be recommended accounts that we’ve found regularly share age-inappropriate content, or if their name or bio suggests the account is inappropriate for teens.
Qualifications & exceptions
Applies to what is recommended to teens.
The first trigger is that the publisher has "found" the account regularly shares age-inappropriate content; the second is the account's name or bio suggesting it is inappropriate for teens.
This item is worded "will not be recommended", without the "generally" used in the introduction to the list.
Not stated here
The page does not define "age-inappropriate content" or say how name or bio is assessed.
People may be told when they are about to engage with an entity that meets the criteria
The document says it may let people know when they are about to engage with an entity that meets any of the above criteria, to help them make informed decisions.
Supporting quotations
We may let people know when they're about to engage with an entity that meets any of the above criteria to help them make informed decisions.
Qualifications & exceptions
Worded "may": a notice is possible, not promised.
The sentence names "an entity"; it says "any of the above criteria".
Not stated here
The page does not say what the notice looks like, when it appears, or whether it also applies to accounts.
Content & IP Rights · Context
A similar set of guidelines applies to recommendations on Instagram
The document says a similar set of these guidelines applies to recommendations on Instagram, and that those guidelines are in the Instagram Help Center. The Instagram guidelines are not supplied here.
Supporting quotations
A similar set of these guidelines applies to recommendations on Instagram.
Qualifications & exceptions
The Instagram guidelines are described as "similar", not identical.
Not stated here
The Instagram guidelines are linked, not supplied, so how they differ from this document is unknown.