Facebook Ad Personal Attributes Policy

What the personal attributes policy actually prohibits

Meta's rule is short, and it is easy to misread. An ad may not assert or imply that you know a personal attribute of the person seeing it. The Transparency Center names fourteen protected categories: race, ethnicity, religion, beliefs, age, sexual orientation or practices, gender identity, disability, physical or mental health (including medical conditions), vulnerable financial status, voting status, membership in a trade union, criminal record, and name. The reasoning Meta gives is plain. An ad that appears to know something private about you can feel intrusive, unsettling, or simply wrong, so the platform asks advertisers to describe the product being sold instead of the audience being sold to.

The word doing the heavy lifting is "imply." Most advertisers assume they are safe because they never state a flat fact about the reader. They ask a question, or address "people who," or lean on the second person, and feel that leaves the attribute unspoken. The policy disagrees. If the natural reading of a line is "this ad is talking to someone who has X," then it asserts the attribute by implication, and that is enough for a rejection. You do not have to say "you have diabetes" to break the rule. "Struggling with your diabetes?" says it for you.

What the policy allows is just as important, and it is easy to lose sight of once you start second-guessing every word on the page. Broad references that are not on the protected list are fine. Calling someone a New Yorker or an American points at where they live, not a private trait. A passing mention of an age group or a gender is acceptable when it is not framed as an assertion aimed at the reader. Names of celebrities and fictional characters are fine. And the word "you" is not banned at all, despite the folklore that says otherwise. "You" only becomes a problem when it is fastened to a protected attribute. "You will love the fit" is fine. "Struggling with your arthritis?" is not, because "your arthritis" claims you have arthritis. The safe version of "you" talks about what the reader will get, never about what the reader is.

This is also why the topic you sell is rarely the real problem. You can advertise a diabetes-friendly cookbook, a debt-consolidation service, a hearing aid, or a dating app for a specific community. Meta expects those businesses to exist and to advertise. The rejection is about the sentence, not the sector. The line that names the product and its outcome sails through; the line that names the reader's condition does not.

The violating patterns, line by line

Almost every personal attributes rejection traces back to one of a handful of copy shapes. Once you can see them, you can catch them in your own drafts before Meta does. The table below pairs the rejected pattern with the attribute it implies and a compliant rewrite that sells the same thing without pointing at the reader.

Two of Meta's own examples make the mechanism unusually clear, and they are worth memorizing because they draw the exact line. "Meet Hispanic men online now!" and "Date Christian singles!" are allowed. "Meet other Black singles near you!" and "Are you Christian?" are not. The difference is not the topic, and it is not even how sensitive the attribute is. It is where the attribute lands. "Date Christian singles" describes who the service is for, as a category of member. "Are you Christian?" asks about the reader. The little words "other" and "are you" flip a compliant line into a violating one, because they move the attribute from the product onto the person scrolling.

The possessive is the other reliable tell. "Your diabetes," "your acne," "your debt," "your depression": each one hands the reader ownership of a private condition. Delete the possessive and describe the thing itself, and the same sentence usually survives. "Support for healthy blood sugar" is a product benefit. "Managing your diabetes" is a claim about the reader. The distance between them is one word, but it is the word the review is built to catch.

Why "for people who..." now gets caught

Advertisers have long known that questions like "Are you overweight?" get rejected, so the workaround that spread was to reframe in the third person: "For people trying to lose weight," "A solution for those living with chronic pain," "Built for anyone struggling with debt." The logic was that describing an audience is different from addressing a reader. In practice it is not, and the policy never treated it as different.

The rule covers implication, and "for people who have X" implies knowledge of the audience just as plainly as "do you have X?" does. When an ad is served to a person and the headline reads "For people managing debt," the natural reading is that the ad believes this person is managing debt. That is an inference about a protected attribute, delivered to an individual, which is exactly what the policy exists to prevent. The third-person grammar changes the sentence but not the implication, and the implication is what fails.

Enforcement has also tightened around this, which is why copy that slipped through a few years ago can be rejected today. Meta's review relies primarily on automated technology, applied to the millions of ads that flow through the system, and it is designed to review all ads before they go live. Automated classifiers reading text at that scale are good at spotting the "for people who" and "do you struggle with" templates because those templates are formulaic. The more your copy sounds like a fill-in-the-blank frame around a condition, the more likely it is to be flagged, whether the frame is a question, a statement, or a third-person label. Reframing the grammar is not a fix. Removing the claim about the reader is.

There is a genuinely compliant version of third person, and it is worth separating from the trap. A real testimonial that describes one named or anonymous customer's experience ("One customer cut her monthly card interest in half in three months") tells a story about a specific person who chose to share it, not an assumption about the stranger seeing the ad. That is allowed, and it is often the strongest creative you can run for a sensitive category, because it demonstrates the outcome without ever labeling the viewer.

The rewrite formulas that pass

Fixing a personal attributes rejection is not a hunt for banned words. It is a shift in who the sentence is about. Four formulas cover almost every case, and each one moves the attribute off the reader and onto the product, the outcome, or a willing narrator.

First, describe the product and the outcome, not the person. This is the workhorse. "Struggling with your diabetes?" becomes "Small daily habits that support healthy blood sugar." "Tired of being overweight?" becomes "A meal plan built around simple, whole-food recipes." The reader recognizes the offer without being told what they are. This is exactly what Meta means when it says ads should focus on the benefits of the product or service.

Second, write in the third person as a real testimonial. "One customer with sensitive skin finally found a routine that did not sting" is a story, not an assumption. The key is that it describes a specific experience someone reported, not a label you are pinning on the viewer. Keep it concrete and attributable rather than a vague stand-in for "people like you."

Third, keep "you" attached to what the reader gets, not to what the reader is. "You have trouble sleeping" asserts a condition. "You will get a wind-down routine you can start tonight" describes a benefit. Same second person, opposite side of the policy line. Whenever you write "you" or "your," check what noun follows it. A product or an outcome is safe; a condition, a status, or an identity is not.

Fourth, ask about a universal desire rather than a private trait. "Want clearer skin?" is a question any person could answer, so it does not single out the reader's condition. "Tired of your acne?" assumes the reader has acne. "Ready to travel more this year?" is fine; "Are you retired and bored?" is not. The test Meta effectively applies is whether the line could reasonably be addressed to any user or only to someone with a specific attribute. Write to the aspiration, which everyone can share, not to the affliction, which only some have.

Here is the same offer run through all four moves so the shift is concrete. A supplement brand starts with the rejected line "Struggling with your joint pain?" The product-focused version is "Support for flexible, comfortable joints." The testimonial version is "One reviewer said the morning stiffness eased within a month." The benefit-anchored "you" version is "You will get a daily formula built around clinically studied ingredients." The universal-desire version is "Want to move more freely this year?" Four compliant lines, one product, and not one of them tells the reader what is wrong with their body.

Health, finance, and identity: the three highest-risk lanes

Three verticals generate the majority of personal attributes rejections, and each carries a second layer of rules on top of the copy policy. Knowing where they overlap saves a lot of guessing.

Health and wellness is the most common. Beyond the copy rule, Meta restricts health and wellness advertising more tightly than most categories: before-and-after imagery, claims that imply an ideal body state, and language that assumes a medical condition all draw scrutiny, which puts a gym or fitness studio advertising body transformations squarely in the firing line. A weight-loss transformation photo paired with "Tired of being overweight?" fails twice over, once for the restricted imagery and once for asserting a physical attribute. The compliant path is to sell the product, the ingredients, the routine, or a realistic and non-dramatized outcome, and to keep the copy off the viewer's body entirely.

Finance is the second. Copy that names a financial hardship ("Drowning in debt?", "Bad credit? No problem", "For people who can't pay their bills") asserts vulnerable financial status. It has to be rewritten around the service ("A debt-consolidation plan that combines your balances into one payment"). Finance ads also usually fall under Meta's special ad categories, a separate framework that exists to reduce discriminatory advertising in credit, employment, housing, and social or political ads. Ads in those categories face targeting restrictions on top of the copy rules: age is locked to 18-65+ with no narrowing, all genders must be included, and detailed-attribute exclusions are not available. So both what you say and who you target are constrained when money is the subject.

Identity is the third: religion, race, sexual orientation, gender identity, and age. The copy rule is the same, but the targeting ground shifted years ago and has not shifted back. On January 19, 2022, Meta removed a large set of detailed-targeting options that referenced sensitive topics, including health causes (options like "Lung cancer awareness" or "World Diabetes Day"), sexual orientation (such as "same-sex marriage" and "LGBT culture"), and religious practices and groups (such as "Catholic Church" and "Jewish holidays"). Meta said it acted in response to concerns from civil rights experts and policymakers about advertisers abusing those options. Campaigns using the removed options kept delivering for roughly eight weeks, until March 17, 2022, and then stopped. The practical result is that you can no longer lean on the attribute in either place: you cannot imply it in the copy, and you cannot target it directly anymore. The way to reach a community is to describe a product that community wants and let broad delivery and strong creative find them.

It is not only the copy: images, targeting, and the landing page

A benefit-focused headline is necessary but not sufficient, because the policy is enforced across the whole ad and its destination. Meta states that its review examines the specific components of an ad, including images, video, text, and targeting information, as well as the ad's associated landing page or other destinations. In other words, a clean headline can still be dragged down by the picture next to it or the page it points to.

The image is the most frequent blind spot. A photo that isolates a medical device on a visibly distressed person, a graphic that singles out a body type, or a visual that reads as "this is you and here is your problem" can assert an attribute the copy carefully avoided. If a stranger glancing at the creative would infer a private condition about the intended viewer, treat the image the same way you treat the text.

The landing page is the second. Advertisers polish the ad and forget the destination, then open the page with "Are you diabetic?" or "For seniors on a fixed income." Meta also requires that the products and services promoted in an ad match those promoted on the landing page, so the review has reason to look past the click. A compliant ad that leads to a non-compliant page is still a non-compliant ad. Read the first screen of your landing page with the same eye you used on the headline, and strip any line that tells the visitor what they are.

Targeting rounds it out. You cannot use audience selection to say what the copy is not allowed to say. Building an ad set around a proxy for a protected attribute, then writing "people like you will love this," recreates the exact implication the policy forbids. Since 2022 the most sensitive targeting options are gone anyway, and for credit, employment, housing, and social or political ads the special ad categories lock the remaining levers. Broad targeting paired with a benefit-first message is now both the compliant path and, in most accounts, the better-performing one.

What to do when an ad is rejected for personal attributes

Start with the notice itself. Meta names the specific policy an ad tripped, so confirm it actually says personal attributes rather than one of the other reasons Facebook ads get rejected, because the fix is different for each. If it is personal attributes, the problem is a claim about the reader somewhere in the text, image, or destination, and your job is to find and remove it.

Rewrite the offending element, do not nudge it. Changing "your" to "the" occasionally works, but more often the whole line has to move from the reader to the product, using one of the four formulas above. Fix every place the implication appears, including the image and the first screen of the landing page, so the resubmission is clean rather than half-corrected. Then resubmit the edited ad or request another review from the Account Quality section. Review is automated first and typically completes within about 24 hours, though it can run longer, and ads can be reviewed again even after they go live.

Two habits protect the account. First, do not repeatedly resubmit the same rejected copy hoping for a different reviewer. A pattern of the same violation is exactly the signal that weighs on account quality, and it can slow future reviews. Change the copy, then resubmit. Second, if a batch of ads was rejected for the same reason, correct the underlying line once and appeal them together rather than one at a time.

Build the timeline in, too, because a rejection is not free. Editing a live ad's text, image, or targeting sends it back through review, so a last-minute fix can pause delivery right when you need it live. Media is not cheap: Gupta Media's tracker put the blended Meta CPM near $8.19 across 2025, and it roughly doubles around peak retail moments, so an ad stuck in re-review during a sale window is lost reach at the priciest time of year. Leave 48 to 72 hours of lead time before any dated launch, and never submit an hour before a promotion opens.

A pre-launch checklist that keeps you compliant

The reliable way to avoid this rejection is to write for it from the first draft, not to fix it after the fact. On a platform whose ads reached an estimated 2.28 billion people as of early 2025, the review is automated and unforgiving of formulaic copy, so the discipline pays off on every ad you ship. Run each draft through five quick checks before you submit.

Read every line and ask one question: does this claim I know something private about the reader? If a sentence only makes sense addressed to someone with a specific condition, status, or identity, it fails. Rewrite it to describe the product or the outcome instead.

Check what follows every "you" and "your." A product or a benefit is safe. A condition, a body state, a financial situation, an age, a faith, or an orientation is not. "You will get" is fine; "you have" and "your condition" are not.

Describe the service's members, not the viewer, when the audience is defined by a protected trait. "Date Christian singles" and "a dating app built for the LGBTQ+ community" describe the product. "Are you Christian?" and "Proud to be gay?" describe the reader.

Look at the image and the first screen of the landing page with the same eye. The policy covers the whole ad and its destination, so a compliant headline over a labeling image, or leading to a "Are you diabetic?" page, still fails.

Drop any targeting that stands in for a protected attribute, and remember the sensitive detailed-targeting options were removed back in 2022. Reach the right people with a product they want and broad delivery, not with a proxy for who they are.

None of this dulls the copy. It usually sharpens it, because a line about the reader's problem is a line about the reader, while a line about the outcome is a line about the sale. "Support for healthy blood sugar" and "move more freely this year" sell the result, which is what the buyer actually wants and what the policy actually rewards. Writing benefit-first from the start, whether by hand or with a tool like AdPlay.ai that drafts on-brand variations, means you clear review and pitch the outcome in the same move, and you spend your launch window running the ad instead of arguing with the reviewer.