Communication Safety is an Apple parental control that inspects photos and videos arriving on, or leaving, a child's iPhone, iPad, Mac, Apple Watch or Apple Vision Pro, and hides anything that appears to contain nudity behind a blur and a warning. As of September 2026 it also intervenes on gore and violent content. The analysis happens on the device, using machine learning, and Apple states that it receives neither the media nor any signal that something was flagged.

The feature exists because explicit images are a recurring lever in the coercion of minors, and because Apple wanted a way to interrupt that exchange without reading encrypted messages. Erik Neuenschwander, Apple's director of user privacy and child safety, described the goal in 2023 as breaking the chain of coercion and influence that makes children susceptible to abuse.

How the checks work on a device

A photo or video lands in a supported app. Before it is displayed, an on-device classifier decides whether it appears to contain nudity. If it does, the content is blurred and the child sees a series of interventions: an explanation, age-appropriate resources and an option to message a trusted adult. Children under 13 are prompted specifically to contact a parent or guardian.

A child who still wants to view the item is asked to confirm and is offered alternatives. Since iOS 18, iPadOS 18, macOS Sequoia and visionOS 2, released in 2024, a child under 13 cannot open flagged content without a parent entering the Screen Time passcode, according to Apple's support documentation. Outgoing media follows the same path: the warning appears before sending, not after.

On iPhone and iPad, Apple lists Messages, AirDrop, Contact Posters in Phone and Contacts, FaceTime calls and video messages, shared photo albums, and the photo picker in some third-party apps. Mac omits AirDrop and FaceTime. Apple Watch covers Messages, Contact Posters and FaceTime video messages. Apple Vision Pro covers Messages, AirDrop and third-party pickers.

Who switches it on

Control sits with the family organiser. The child must be signed in with a child Apple Account inside a Family Sharinggroup, and the setting lives under Screen Time.

Defaults have moved several times. Apple's support page now states that the feature is on by default for children under 18 on iPhone, iPad, Mac and Apple Watch running current software, and on by default for 13 to 17 year olds on Apple Vision Pro. Parents can still turn it off.

Adults have a separate, opt-in equivalent called Sensitive Content Warning, found under Privacy and Security settings.

The developer layer

Apple opened the classifier to other apps in 2023 through the SensitiveContentAnalysis framework, available from iOS 17 and macOS 14. An app must hold the com.apple.developer.sensitivecontentanalysis.client entitlement with the value analysis. Without it, calls fail to return positive results.

The central object, SCSensitivityAnalyzer, reports an analysisPolicy with three values. Disabled means neither user setting is active, the entitlement is missing, or the user switched the check off for that app. simpleInterventions reflects Sensitive Content Warning and calls for discreet treatment, such as blurring in place. descriptiveInterventions reflects Communication Safety and calls for overt, explanatory treatment.

Apple's developer licence adds a hard limit: an app may not transmit off the device any information about whether the framework identified an image or video as containing nudity. In July 2026 Apple listed the framework as newly capable of detecting violent imagery and extending its real-time check into video calls.

Origin and evolution

Apple announced the feature on August 5, 2021, under the banner Expanded Protections for Children, alongside a separate plan to match iCloud Photos uploads against hashes of known child sexual abuse material (CSAM). The first design included an option to notify parents of children under 13 when a flagged image was viewed or sent.

On August 19, 2021, more than 90 organisations coordinated by the Center for Democracy and Technology (CDT) urged Apple to drop both plans. The letter argued that an abusive adult could be the family organiser and that LGBTQ+ youths with unsympathetic parents were particularly at risk from notifications.

Apple removed parental notification before launch, according to TechCrunch, and CDT welcomed the change on November 11, 2021. The feature shipped with iOS 15.2 on December 13, 2021, in the United States, covering still images in Messages only and requiring parents to opt in. Expansion to the United Kingdom, Canada, Australia and New Zealand followed in 2022. The CSAM matching plan, paused in September 2021, was confirmed abandoned in December 2022.

iOS 17 in 2023 was the largest step. Video detection arrived, along with AirDrop, Contact Posters, FaceTime video messages and the system photo picker. According to MacRumors, the feature also went worldwide and became default for under-13 accounts in Family Sharing.

On October 24, 2024, Apple added a reporting option in the iOS 18.2 beta for children in Australia. A report bundles the flagged media, messages sent immediately before and after, and contact details for both accounts; Apple said it could disable the sender's account or refer the matter to law enforcement. Australia's eSafety Commissioner welcomed the step, noting it came nearly two years after its own transparency reporting found such basic measures missing.

In June 2025 Apple said iOS 26 would intervene when nudity appears in live FaceTime video calls and blur it in Shared Albums, and would apply Communication Safety by default to teens aged 13 to 17, according to 9to5Mac.

Why it matters for marketers and publishers

Communication Safety is one part of a device-level child safety stack that increasingly shapes commercial reach. The Declared Age Range framework returns age bands to apps, the five-tier App Store rating system introduced in July 2025 governs discovery, and Apple Ads, according to Apple, serves no ads to children under 13 and no personalised ads to older minors.

For publishers of user-generated content, the placement matters most. A filter that intercepts media before display sits upstream of any moderation decision a platform makes on its own servers. Messaging and social apps adopting the framework inherit Apple's classifier, its thresholds and its silence about both.

Outside the device, age assurance duties now run from Texas to the United Kingdom, where the Online Safety Act received Royal Assent on October 26, 2023. In the European Union, the temporary legal basis allowing platforms to scan voluntarily for abuse material expired on April 3, 2026, which leaves on-device, user-facing tools as one of the few mechanisms operating without that derogation.

Limitations, criticisms and disputes

Accuracy is the first gap. Apple publishes no false positive or false negative rates. The 2021 coalition letter warned that such classifiers are prone to flag art, health information and educational material. Google, describing its own equivalent, concedes the check sometimes misses nudity and sometimes flags images that contain none.

Scope is the second. Protection applies inside Apple apps and inside third-party apps that choose to integrate; any other app is outside it. Settings depend on Family Sharing and on every device in the group being updated, and Apple states some features may not reach every region, language or model without naming which.

Child safety groups argue the feature is an insufficient substitute for detection. Heat Initiative campaigned in 2023 for Apple to detect and report CSAM in iCloud. In reply, Neuenschwander wrote that such scanning was "not practically possible to implement without ultimately imperiling the security and privacy" of users. The UK's National Society for the Prevention of Cruelty to Children (NSPCC) found Apple implicated in 337 recorded offences in England and Wales between April 2022 and March 2023, while Apple made 267 reports worldwide to the National Center for Missing and Exploited Children in 2023, according to Tom's Guide.

Privacy advocates raise the opposite concern. The 2021 letter warned that detection built for one category could later be pointed at others. Apple has now widened the classifier from nudity to violence on its own initiative, and the Australian reporting flow sends flagged media to Apple when a child chooses to report. Neither contradicts Apple's on-device claim, but both show the boundary moving.

PPC Land noted that Apple has not said how the violence classifier works, where inference runs or what latency the FaceTime path adds.

Not the same as

CSAM detection matched uploads against a database of known abuse images and was abandoned. Communication Safety classifies any image, known or not, and reports nothing by default.

Sensitive Content Warning is the adult, opt-in version and maps to the simpleInterventions policy rather than the descriptive one.

Communication Limits controls who a child may contact, not what the content contains.

Sensitive content warnings in Google Messages perform a similar on-device job through Android System SafetyCore, launched widely in August 2025. Google's version currently covers images only, is on by default for supervised users and signed-in teens, and is off by default for adults.

Recent developments

Apple previewed the violence expansion on June 8, 2026, describing Communication Safety as already on by default for users under 18. The change shipped on September 14, 2026, with iOS 27, iPadOS 27 and macOS 27, alongside Ask to Browse, category Time Allowances and parental approval of new contacts. Apple's announcement states that protections now block gore and violent content in shared images, videos and live FaceTime calls.

The same month, answers to Apple's social media rating questions became mandatory for App Store submissions. Every device in a Family Sharing group must run the new software before the updated controls function at all.

Timeline

  • August 5, 2021: Apple announces Expanded Protections for Children, including Communication Safety in Messages and CSAM detection for iCloud Photos
  • August 19, 2021: More than 90 organisations coordinated by CDT publish an open letter opposing both plans
  • September 2021: Apple pauses the CSAM detection project
  • November 2021: Parental notification for under-13 accounts is removed from the design before launch
  • December 13, 2021: Communication Safety ships in the United States with iOS 15.2 and macOS 12.1, covering images in Messages
  • 2022: Rollout extends to the United Kingdom, Canada, Australia and New Zealand
  • December 2022: Apple confirms the CSAM detection project is abandoned
  • 2023: iOS 17 and macOS 14 add video detection, AirDrop, Contact Posters, FaceTime video messages, the photo picker, the SensitiveContentAnalysis framework and default activation for under-13 accounts
  • August 2023: Apple responds in writing to Heat Initiative's campaign
  • 2024: iOS 18 requires a Screen Time passcode before under-13 users can view flagged content
  • October 24, 2024: iOS 18.2 beta adds reporting to Apple for children in Australia
  • February 2025: Apple publishes the Helping Protect Kids Online white paper
  • June 2025: Apple announces FaceTime video call intervention, Shared Albums blurring and default protection for teens aged 13 to 17
  • August 2025: Google's comparable Sensitive Content Warnings roll out widely in Google Messages
  • September 2025: iOS 26 ships with the expanded teen defaults
  • June 8, 2026: Apple previews gore and violence detection at WWDC26
  • July 2026: Apple lists SensitiveContentAnalysis as able to detect violent imagery
  • September 14, 2026: iOS 27, iPadOS 27 and macOS 27 ship with violence detection, including live FaceTime calls

Summary

Who: Apple builds and operates Communication Safety. Parents control it through Family Sharing, children under 18 are protected by default on most Apple hardware, and third-party developers can plug into the same classifier through SensitiveContentAnalysis.

What: An on-device machine learning check that blurs photos and videos appearing to contain nudity, and since September 2026 gore and violence, then presents warnings, resources and options to seek help before a child views or sends them.

When: Announced on August 5, 2021, launched on December 13, 2021 with iOS 15.2, expanded to video and more apps in 2023, extended to live FaceTime calls with iOS 26 in 2025 and to violent content on September 14, 2026.

Where: Messages, AirDrop, FaceTime, Contact Posters, Shared Albums and participating third-party apps across iPhone, iPad, Mac, Apple Watch and Apple Vision Pro, with availability varying by region.

Why: Apple chose it as a privacy-preserving alternative to scanning stored content, intervening at the moment of exchange without giving Apple access to messages. Child safety groups consider it insufficient, privacy advocates watch its scope, and advertisers and publishers now operate inside the device-level controls it belongs to.