Ad Protection
Ad Protection continuously watches the Meta ad accounts linked to your brands. If someone advertises from your ad account without permission - for example, a Facebook profile with access was hacked, or an unknown partner has access -, ad protection spots the foreign ads within minutes, alerts you by email, and can even step in.
How It Works
Ad protection continuously goes through the running ads of the ad account and checks the ads that were started or changed.
Ads created in Infinite∞Ad and boosts of the brand's own posts are trusted, so they are not scored. Every other ad is compared by Infinite∞Ai to the usual ads of the ad account, its content is evaluated too, and it gets a risk score. You can see why an ad got a high score on the ad's details. Based on the score, an ad can be suspicious or worth a look (see: Ad States). For a suspicious ad, an incident is opened, and if automatic pausing is on, the ad is paused right away.
Ad protection also watches the spending limit of the ad account, because an intruder often raises or removes it first. If the limit is raised or removed in Meta, an incident is opened even without a suspicious ad, and you get an email alert. (A spending limit change made in Infinite∞Ad does not trigger an alert.)
First Check
When turned on, ad protection checks all running ads of the ad account once. The first check only reports, it pauses nothing - this is when the system learns the usual ads of the ad account. If it finds suspicious ads, it emails you the result, and you find the suspicious ads among the Reviewed Ads.
If an ad account was not checked for a longer time (for example, monitoring was turned off, or its brand was inactive), the next check is again a report-only check.
Guarding Paused Ads
The ads paused by ad protection are guarded: if someone restarts them, they are paused again. If an ad is restarted several times, its ad set is paused too. If every running ad of a campaign is suspicious, the campaign is paused as well. A guarded ad cannot be restarted from Infinite∞Ad either until someone marks it as safe.
Turning On and Settings
You can turn on and set up ad protection in two places:
- when creating or editing a brand, in the Ad account protection section;
- on the Overview tab of Ad Protection, with the Settings button of the ad account.
If several brands use the same Meta ad account, the ad protection settings are shared: what you change at one brand applies to the others too. In this case, the settings also show which other brands they apply to.
Ad protection can be on for a Meta ad account in only one organization at a time. If another organization already protects the ad account, you see On in another organization, and monitoring cannot be turned on.
Settings
- Monitoring: turns ad protection on or off for the ad account.
- Automatic pausing: when on, suspicious ads are paused right away, and paused again if someone restarts them. When off, you only get alerts. If you turn it on later, the ads flagged as suspicious before that and still running are paused at the next check.
- Sensitivity: from what risk score an ad counts as suspicious. With a stricter setting, the system alerts sooner, but you have more ads to review.
- Send alerts to: you can choose members of the organization and add additional email addresses. At least one recipient is required.
Known Domains
Ads linking to known domains don't count as suspicious for their link. Infinite∞Ad collects them automatically from the websites of the brands, the earlier ads, the posts, the automations and the ads marked as safe. For each domain you can see where it was found.
You can also add domains by hand and remove any of them. A removed domain is not added again automatically, only if you restore it or add it again by hand.
Link shorteners and shared hosting (for example bit.ly or linktr.ee) are never added to the list automatically, because they hide the real landing page. If you use them anyway, add them by hand.
The Ad Protection Page
In the navigation bar, a red counter next to the Ad protection menu item shows the number of open incidents. The page has three tabs.
Overview
- Summary tiles: the number of open incidents (if there are none, the number of days since the last incident), the ads waiting for your decision, and the ads checked and paused in the selected period.
- Ads waiting for your decision: the ads that may still be running and nobody has decided about yet.
- Ad accounts: the protection state of the ad accounts, the time of the last check, and the Settings button. If there is an open incident, you can open it directly from here.
- Latest events: a chronological list of what ad protection did.
Ad Account States
| State | Meaning |
|---|---|
| Protected | Ad protection watches the ad account, and there is nothing for you to do. |
| Needs attention | There is an ad waiting for a decision, or the last check failed. A failed check is retried automatically. |
| Incident open | A suspicious ad or a spending limit change was found in the ad account, and the incident is not resolved yet. |
| Waiting for the first check | Monitoring is on, but the first check has not run yet. |
| Off | Ad protection does not watch the ad account. |
| On in another organization | Another organization protects the ad account. |
Incidents
The incidents in chronological order, with a brand and a status filter (open or resolved). Read more about incidents in the Incidents section.
Reviewed Ads
The ads that ad protection checked and flagged, or that someone in the organization marked as safe. You can filter the list by brand and state (Waiting for your decision, Paused, Safe), and search for an ad, campaign or domain.
Ad States
| State | Meaning |
|---|---|
| Suspicious ad, still running | The ad is suspicious but not paused - for example, because automatic pausing is off, or the first check found it. You need to decide about it. |
| Ad protection paused this ad | The ad was suspicious, and the system paused it. If someone restarts it, it is paused again. |
| Ad protection could not pause this ad | Meta did not let the ad be paused. The system retries, but it is worth pausing it by hand in Ads Manager. |
| Ad worth a look | The ad is not suspicious, but unusual: it is worth a look. |
| Ad waiting for a review | Infinite∞Ai could not check the ad. Review it and decide whether it is safe. |
| Ad marked as safe | Someone in the team marked it as safe. |
| Checked ad | The check found nothing suspicious in it. |
Ad Details
Clicking on an ad opens its details: the ad preview, where it runs (brand, ad account, campaign), its budget, its landing page and the targeted countries. The Reasoning part shows the risk score, the signals found and the explanation of Infinite∞Ai, and the History part shows the events and incidents of the ad. With the Open in Ads Manager link you can also view the ad in Meta.
At the bottom of the details you can choose from two actions:
- Pause: the ad is paused in Meta right away, and if someone restarts it, ad protection pauses it again.
- Mark as safe: ad protection trusts the ad and stops guarding it. If the ad links to a website, the domain of that website is added to the known domains. Marking a paused ad as safe does not restart it: if you want it to run, restart it in Ads Manager or in Infinite∞Ad.
Incidents
An incident opens when ad protection finds a suspicious ad in an ad account, or when the spending limit of the ad account is raised or removed in Meta. An ad account can have one open incident at a time: while it is open, new suspicious ads are added to it.
The name of an incident is a number and a title (for example #7 Investment scam from an unknown page). The title is given by Infinite∞Ai based on the ads of the incident. With the Rename command in the "…" menu of the header you can give it your own title; if you leave it empty, Infinite∞Ai names it again.
The Incident Page
- What to do now - the list of tasks according to the state of the incident: pausing the suspicious ads, checking who can reach the ad account in Meta Business settings (with direct links), setting the spending limit back, marking ads as safe, and finally resolving the incident.
- Affected ads - the ads of the incident with their states.
- Timeline - what happened in the incident, to the minute: detection, pausing, alerts, the steps of the team.
Pausing All Unknown Ads
If you think someone unauthorized has access to the ad account, on the incident page you can pause every ad of the ad account that Infinite∞Ad does not know with a single button. Only the ads created in Infinite∞Ad, the boosts of the brand's own posts and the ads marked as safe keep running. The paused ads are guarded.
This also pauses the ads made in Ads Manager by your colleagues or your agency. Nothing is restarted automatically: you can mark your own ads as safe, and then restart them.
Resolving and Reopening
Incidents are not resolved automatically: once the ad account is secured, someone in the organization has to resolve the incident with the Resolve button in the header. If every ad of the incident is paused or marked as safe, a highlighted bar offers resolving it, and the list shows Can be resolved. Before resolving, check who can reach the ad account.
If someone restarted a paused ad, a warning bar appears: the intruder may still have access to the ad account. In this case, check the access first, and resolve the incident only after that.
If a paused ad is restarted within 7 days after the incident was resolved, the incident reopens by itself. New suspicious ads open a new incident. You can also reopen a resolved incident by hand with the Reopen button, if the ad account has no other open incident.
Notifications
The alerts are sent by email to the recipients set in the settings:
- Incident opened: right away, with the affected ads.
- Incident updated: when a new suspicious ad is added to the incident, or someone restarts a paused ad. When ads come in quick succession, the system sends them together.
- Incident resolved: who resolved it, and a warning if a suspicious ad was left running.
- Spending limit change: when the spending limit of the ad account is raised or removed in Meta.
- First check: the result, if the check found a suspicious ad.
While an incident is open, a notification also appears in Infinite∞Ad, which disappears when the incident is resolved.