Apple Will Start Scanning Publicly Shared iCloud Links for "Harmful Content"

Apple is changing how it handles one specific corner of iCloud sharing, and the shift comes with a quiet but significant trade-off: lose your encryption, gain a content check.

Starting with iOS 27.3 and the matching updates for iPadOS, macOS, and visionOS, any iCloud item you share using the "Anyone with the Link" setting will become accessible to Apple so the company can scan it for harmful content. That's according to an Apple support document first reported on by 9to5Mac. The company spells it out plainly: content shared that way "is made accessible to Apple to check for harmful content."

There's a catch tied directly to that access. Sharing a file through "Anyone with the Link" already meant it was no longer end-to-end encrypted, and that remains true, now paired explicitly with the scanning capability. In other words, the moment you open a link to the public rather than named recipients, you're opting out of Apple's strongest privacy protections and opting into a review process.

What counts as "harmful content" isn't defined anywhere in Apple's documentation. The company doesn't specify categories, thresholds, or what triggers a flag. What it does spell out is what happens after something gets flagged: Apple says it reserves the right to move, refuse, modify, remove, or delete flagged material. Beyond that, the company can block or disable accounts, stop someone from creating a new one, and, "if necessary," bring in law enforcement.

Why now

Apple doesn't frame this as a response to any single incident, but the timing lines up with years of pressure the company has faced over child sexual abuse material, or CSAM, stored in iCloud. Protesters have repeatedly called on Apple to do more to detect and remove it, and this new scanning policy for publicly shared links looks like a direct answer to that pressure, even though Apple's own documentation stays silent on CSAM specifically.

It's a notably narrower approach than the one Apple floated and then abandoned a few years back. That earlier plan involved scanning photos on-device before they uploaded to iCloud, a proposal that drew fierce backlash from privacy advocates and security researchers who warned it created a backdoor that could be expanded or abused. Apple shelved it. This time, the scope is tighter and the trigger is user-initiated: the scanning only applies when someone actively chooses to make an item accessible via a public link, not to iCloud content broadly or to files shared privately with specific people.

That distinction matters. It lets Apple frame this as checking content once it's effectively public, rather than monitoring private communications or personal storage. Whether that framing holds up to scrutiny is a separate question, but structurally it's a different kind of intervention than the on-device scanning Apple walked away from.

What it means for users

Practically, this gives anyone using iCloud sharing a clear line to think about. Share something privately, with specific people, and encryption protections stay intact, as they already did. Flip that same item to "Anyone with the Link" so it's publicly accessible, and you've now handed Apple visibility into it, along with the company's self-granted authority to act on what it finds, up to and including deleting the content or disabling your account.

Apple hasn't said how the scanning technically works, what tools or models are involved, or how false positives might get handled. The support document is procedural rather than technical: it tells users what will happen and what Apple can do about it, not how the detection itself functions.

Apple kicked off the first developer betas for the 27.3 update, and the company has said it's targeting a December release for iOS 27.3 across its platforms. That gives users a short runway before the new policy actually ships, and a reason to pay closer attention to which sharing setting they pick the next time they send someone an iCloud link.