Communication Safety on Apple devices and what it scans

Communication Safety is an Apple feature that blurs images containing nudity in supported surfaces, on the child’s device [source: apple-communication-safety].

It is not a parent-notification feature. It does not send the image to Apple, and it does not tell the parent what the child saw [source: apple-communication-safety-de].

That sentence is the whole article in one line. Everything below is the long version, separating two questions that often get tangled:

• What does the feature do? Detection, blurring, and the child-facing safety note.
• What does the parent see? Nothing, by design, and the privacy reasoning behind that.

Definition

Communication Safety is a feature Apple ships on iPhone, iPad, Mac, and Apple Watch. It analyses images received in Messages, AirDrop, the system Photos picker when sending images, and FaceTime video [source: apple-communication-safety].

When the analysis detects nudity, the image is blurred before the child sees it. The child is presented with a safety interstitial that explains what is happening and offers a way to view the image, talk to someone they trust, or block the contact [source: apple-communication-safety].

The analysis runs on the device. Apple documents this explicitly: the images are not sent to Apple, and Apple does not get access to them [source: apple-communication-safety-de].

For accounts set up as child accounts in a Family Sharing group, Communication Safety is on by default and cannot be turned off by the child [source: apple-communication-safety] [source: apple-family-kids]. For accounts set up as adults, the same feature is called “Sensitive Content Warning” and is opt-in [source: apple-communication-safety].

Scope

This article is about the child side of the feature. It covers:

• Where the feature runs. The supported surfaces, the supported operating system versions, and the language-version caveats.
• What the feature does to an image. Detection, blurring, the safety interstitial.
• What the parent sees. The privacy contract Apple publishes, and the parent’s deliberate absence from the loop.
• What the feature does not do. The cases the feature does not cover.

It does not cover:

• How to set up a child Apple Account. That is documented in the Set up Screen Time on iPhone article and in Apple’s “Create an Apple Account for your child” page [source: apple-family-kids].
• Other family communication features. Communication Limits and Screen Time’s wider surface area live in the Screen Time container, not in Communication Safety.
• Non-Apple platforms. Android, Google Messages, WhatsApp, and Signal each have their own image-detection behaviour. The wiki covers them in app-profile articles.
• The detection model. Apple does not publish the model or its training data. The article cites what Apple does publish, not what is in third-party reporting.

Where the feature runs

The supported surfaces, as Apple documents them [source: apple-communication-safety] [source: apple-communication-safety-de]:

• Messages. Incoming and outgoing images.
• AirDrop. Incoming images in the AirDrop preview.
• Photos picker. When the child selects an image to send from the system picker.
• FaceTime video. Live video previews are blurred when nudity is detected; the participant on the other end is not informed.
• Contact Posters in the Phone app. The poster images shown when receiving a call.
• Third-party apps that opt in. Apple publishes an API that lets third-party apps integrate Communication Safety. The integration is opt-in for the app developer, not automatic [source: apple-communication-safety].

The list above is what Apple publishes. A child who uses a third-party app that has not integrated the API is outside the feature’s coverage.

The minimum operating system version is iOS 15.2 / iPadOS 15.2 for the original Messages surface, and iOS 17 / iPadOS 17 for the broader app coverage [source: apple-communication-safety-de]. The under-13 viewing gate (the child cannot disable the feature) is part of the iOS 18 family configuration [source: apple-communication-safety].

The screen reference for the parent-side entry, in case a parent needs to confirm the toggle state, is apple.screentime.communication_safety.

What happens when an image is flagged

The sequence is mechanical. Apple documents the flow [source: apple-communication-safety] [source: apple-communication-safety-de]:

1. The image arrives in a supported surface. A message, an AirDrop preview, a FaceTime frame, or a third-party app that has integrated the API.
2. The on-device model evaluates the image. The evaluation is local. It does not depend on a network round trip.
3. If the model flags the image, the device blurs it. The child sees a blurred version with the safety interstitial overlaid.
4. The child chooses what to do. The interstitial offers three actions: view the image anyway, message a trusted contact, or block the sender [source: apple-communication-safety].
5. No notification is sent to the parent. Apple documents this explicitly. The parent does not receive an alert, a log entry, or a report [source: apple-communication-safety-de].

The single sentence version: the feature replaces an image the child would have seen with an interstitial the child can act on. The parent is not in the loop.

What the parent sees

Nothing about the specific image. The parent sees what any other device sees: that Communication Safety is on.

Apple documents the privacy contract as a property of the feature, not as a configuration toggle [source: apple-communication-safety-de]:

• The images are not sent to Apple. The detection model runs on the device. No image leaves the device for analysis [source: apple-communication-safety].
• Apple does not get access to the analysis result. The result is a local decision about whether to blur [source: apple-communication-safety-de].
• No notification is sent to the parent. The family organiser does not see a log of blurred images, a count of interstitials shown, or a flag when a child chose to view an image [source: apple-communication-safety].

The parent who wants visibility into what the child saw is looking for a feature Communication Safety is not. The Screen Time report shows what apps were used; it does not show what was viewed inside an app [source: apple-screen-time].

A separate family-organiser control, Communication Limits, governs who the child can talk to (or be talked to by). It is a different feature, on a different setting, in the same Screen Time container [source: apple-screen-time].

Where detection runs and where it does not

The list below separates surfaces the feature covers from surfaces it does not.

Covered by Communication Safety on iOS 18 / iPadOS 18 [source: apple-communication-safety] [source: apple-communication-safety-de]:

• Apple Messages
• AirDrop
• FaceTime
• Contact Posters in the Phone app
• The system Photos picker when sending images
• Third-party apps that have integrated the API

Not covered, or covered only partially:

• Web pages. Safari does not integrate Communication Safety as of iOS 18. The web content filter inside Content & Privacy Restrictions is a separate layer, and it is rated by domain, not by image content [source: apple-screen-time].
• Third-party messaging apps that have not integrated the API. WhatsApp, Signal, Telegram, and most chat apps do not use the Communication Safety API. The detection does not run on their surfaces.
• Email. Apple’s Mail app does not currently integrate the detection API. The list above is what Apple publishes [source: apple-communication-safety].
• Video on the web. Streaming and embedded video fall outside the supported surfaces.
• Photos already on the device. Communication Safety blurs images at receipt. It does not retroactively scan the photo library.
• Devices not running the supported OS versions. An iPad on iPadOS 16 is not covered.

A child who moves from Apple Messages to a third-party messaging app, or to webmail, has stepped out of the feature’s coverage. The detection follows the surface, not the conversation.

Why this design

Apple publishes the privacy contract in two places: the feature support page and the family sharing page for kids and teens [source: apple-communication-safety] [source: apple-family-kids].

The reasoning, in plain language:

• Detection on the device means Apple never sees the image. This is the difference between an on-device classifier and a server-side one. It is the same architectural choice Apple’s other privacy features use, including Private Relay and Hide My Email.
• No parent notification means the child is the decision-maker. This is the design that distinguishes Communication Safety from surveillance. The child sees the interstitial, the child chooses. The parent is deliberately not in the loop [source: apple-communication-safety-de].
• The child account makes it default-on. For accounts set up as child accounts under Family Sharing, the feature is on by default and cannot be turned off by the child [source: apple-family-kids]. For adult accounts, the equivalent Sensitive Content Warning is opt-in.

The third point is the one parents sometimes miss. The feature is on for a child Apple Account before the parent does anything. The configuration is “did you set up a child Apple Account”, not “did you turn on Communication Safety” [source: apple-family-kids].

What the feature does not do

A short list to set expectations:

• It does not notify the parent. The parent does not see a log of blurred images or a count of interstitials shown [source: apple-communication-safety-de].
• It does not block the conversation. The child can choose to view the image, message a trusted contact, or block the sender. The feature does not stop the conversation [source: apple-communication-safety].
• It does not detect text. The classifier evaluates images. It does not evaluate message text.
• It does not retroactively scan the photo library. Images already on the device, or received before iOS 18, are not re-evaluated.
• It does not extend to non-Apple platforms. A child who moves to a third-party messaging app on the same iPhone has stepped out of the feature’s coverage.
• It does not replace Communication Limits. Communication Limits controls who the child can talk to. Communication Safety blurs what the child sees. The two are independent and configurable separately [source: apple-screen-time].

The single sentence version: the feature replaces an image with an interstitial, locally. It does not tell the parent anything about the conversation.

Talking about it

A feature that the child experiences alone, and the parent knows about only from documentation, is a feature that needs to be explained. Three habits help [source: apple-family-kids] [source: consumer-ftc-parental-controls]:

• Tell the child the feature exists. A child who sees an interstitial they do not understand may panic or assume something is broken. The conversation about what the feature does is the conversation about what the child does next [source: apple-communication-safety].
• Tell the child what the feature does not do. It does not send the image to Apple. It does not send a notification to the parent. It does not block the conversation. The child who knows this also knows the conversation is not being watched in the way they might fear [source: apple-communication-safety-de].
• Re-decide when the child moves to a new app or device. The feature follows the surface, not the conversation. A child who moves from Apple Messages to a third-party chat on the same iPhone has moved out of coverage. The conversation is the same; the detection is not [source: apple-communication-safety].

The point of the feature is not to police the conversation. It is to give the child a pause at the moment an image arrives. The pause is the rule.

Sources

• apple-communication-safety — Apple Support, “About Communication Safety on your child’s Apple device” (EN, retrieved 2026-03-24). The supported surfaces, the on-device analysis, and the privacy contract.
• apple-communication-safety-de — Apple Support, “Informationen zu einer sicheren Kommunikation auf dem Apple-Gerät deines Kindes” (DE, retrieved 2026-08-16). The primary German localization and the parent-notification statement.
• apple-screen-time — Apple Support, “Use Screen Time to manage your child’s iPhone or iPad” (EN, retrieved 2026-07-01). The wider Screen Time container, including Communication Limits.
• apple-screen-time-de-2 — Apple Support, “Bildschirmzeit für ein Familienmitglied einrichten” (DE). The German UI strings for the Screen Time container.
• apple-family-kids — Apple Support, “Family Sharing overview for kids and teens” (EN, retrieved 2026-08-16). The default-on child-account behaviour.
• apple-family-kids-de — Apple Support, German localization of the family sharing kids page.
• apple-family-sharing — Apple Support, “How Family Sharing works” (EN, retrieved 2026-07-29). The family-side context.
• consumer-ftc-parental-controls — U.S. Federal Trade Commission, parental-controls consumer guidance. The periodic-re-review recommendation.

[source: apple-communication-safety] [source: apple-communication-safety-de] [source: apple-screen-time] [source: apple-screen-time-de-2] [source: apple-family-kids] [source: apple-family-kids-de] [source: apple-family-sharing] [source: consumer-ftc-parental-controls]