You search through Meta’s policies for “comparative advertising” and you cannot find it. That is the first reason this question creates so much confusion: the ban everyone assumes exists simply does not.
What does exist is another set of rules that produces much the same practical effect through different mechanisms. Understanding what those rules are completely changes the way you write your creative.
Meta does not ban comparison. It bans confusion.
Meta’s Advertising Standards state that ads cannot contain material that violates third-party intellectual property rights, including trademarks. So far, nothing surprising.
The part that really matters comes next. Ads must not confuse people about the origin, sponsorship, or affiliation of the goods and services being advertised.
This is a criterion, not a ban on mentioning a competitor. It means the problem does not arise because you mention a competitor, but because your ad could make the person seeing it think that competitor has something to do with you. That they authorized you. That you are partners. That the product belongs to them.
Meta itself makes clear that it cannot know with certainty whether an ad infringes someone’s rights without a report from the rights holder, but that it may still take action when there are indications of potential infringement. This is where the behavior everyone knows comes from: automated review is cautious, and that caution lands on you in the form of a rejection.
There is also the actual reporting channel. The trademark owner can report the ad, and established brands have access to dedicated tools for detecting unauthorized uses of their trademarks across ads at scale. If you are advertising against a large, organized player, assume your ad will be seen.
Why your comparative ad gets rejected even when it is lawful
Here is the operational point that makes up the entire reason for this article.
Meta’s approval is not a judgment of legality, and Meta’s rejection is not a judgment of illegality. They are two systems that do not speak to each other.
You can have a comparative ad that complies with all the requirements of Italian Legislative Decree 145/2007, uses verifiable data, and contains no disparagement, yet still have it rejected because the competitor’s packaging appears for two seconds in the video and the classifier interprets it as a potential trademark violation. You have not lost a lawsuit. You have lost a review.
And the exact opposite can happen, which is the dangerous case. The ad is approved, runs for three weeks, and converts extremely well. Then the competitor realizes that the claim could not be substantiated. At that point, the fact that Meta approved it is of no use to you, in front of anyone.
So: the policy is your distribution problem, while the law is your risk problem. They need to be managed separately, with separate controls.
Where the risk really concentrates
In the video and the image, much more than in the text.
You write and control the text. The creative, however, contains a thousand things you may not have evaluated as a “mention of the competitor”: a logo in the background, an interface screenshot, a label captured on camera, a visual comparison between two packages, a frame from one of their ads. Each of these is, to the system, use of someone else’s trademark.
The second hot spot is side-by-side visual comparisons. Two columns, their logo on the left and yours on the right. It is the most effective form of comparative advertising and also the one that triggers the most reports, because it is the most visible and easiest for anyone to recognize.
The third is a testimonial who mentions the competitor while speaking. Spoken language is not transcribed and evaluated in the same way as text, so it may sometimes get through. But the legal risk remains exactly the same, and in the event of a dispute, the video is the evidence.
How to reword a comparative claim so it gets approved
The approach that works best is to shift the comparison from the name to the characteristic.
Instead of naming the competitor, name the market behavior. “Most practice management software for dentists requires an annual contract” is a statement about the market, not an identifiable company. It is not comparative advertising, it does not use a trademark, and it communicates almost the same thing.
But be careful, because there is a trap hidden here. If there are only two operators in your market, the generic statement implicitly identifies the competitor and you are back within the rules governing comparative advertising. Generality may protect you from the platform’s policy, but not necessarily from the law.
The second approach is to move the comparison to the landing page. The ad brings in traffic without naming anyone, while the landing page makes the comparison using data, sources, and dates. You move the sensitive content to an asset you control, which no automated classifier has to approve and which you can correct in five minutes. It is also the approach that makes that content indexable and citable, which is worth far more than the creative itself.
The third is to anchor the claim to data rather than judgment. “It costs less” is a judgment. “In our public price list dated August 12, 2026, the equivalent plan costs X” is a fact. Facts are harder to challenge, easier to substantiate, and, interestingly, convert better because specificity is always more credible than superiority.
Before uploading the creative
Watch the video frame by frame and ask yourself whether a trademark that is not yours appears in any form.
Make sure no element could suggest a relationship between you and the competitor, including phrases such as “official alternative” or “compatible with” when used ambiguously.
Make sure every objective claim has documented, dated evidence behind it, referring to the current version of the other company’s product.
And keep an alternative version of the creative without recognizable references ready to upload. If the rejection comes, you do not want to rewrite the campaign: you simply want to replace the asset and start again.
Content for informational purposes only. It does not replace a legal assessment of a specific case.



