Apple Reference Image in iOS 27 Beta Tackles Deepfakes With Photo Authentication System

Apple could help you prove your iPhone photos aren't deepfakes

Apple is seemingly developing an iOS feature that can verify when a photograph was taken using an iPhone camera. 9to5Mac reports that the iOS 27 beta 5 includes code references for an "Apple Reference Image" system that can embed provenance metadata into iPhone photographs at the point of capture -- enabling users to prove where the photo originated, and that it isn't AI fakery. Apple Reference Image isn't currently live, but a privacy disclosure in the iOS beta notes that the feature will be off by default. When (or if) it does roll out, 9to5Mac says it can be enabled via Settings > Camera > Reference Image > Reference Mode, and that only photographs taken using a new "Reference" option in the iPhone Camera app will include the provenance information required for authentication. MacRumors has more details about the actual authentication process, which doesn't occur automatically and instead requires users to tap a Reference badge on the photograph. This reportedly sends the raw image and its embedded provenance data -- including sensor signatures, capture time frame, and unique hardware identifiers -- to Apple's Private Cloud Compute servers for verification, and then returns the authenticated version with a uniquely assigned ID. Apple doesn't access the raw photo itself during this process, but it may receive sensor data that allows it to prevent images connected to compromised sensors from being authenticated, or retroactively revoke prior authentication on images associated with the sensor. Authenticated photos can be viewed on an iPhone, iPad, or Mac, according to MacRumors. By this description, Apple Reference Image works similarly to the C2PA Content Credentials provenance standard, which can be used to trace how and where an image was made, and if generative AI tools were used to make or manipulate it. Camera companies including Canon, Nikon, Sony, FujiFilm, and Leica have been gradually introducing it into their hardware, and Google's Pixel 10 phone cameras also support the standard, but Apple has notably avoided adopting it. C2PA isn't known for its reliability, however, so perhaps Apple believes its own system will be a better, more robust solution. Introducing a provenance system for iPhone photography also suggests Apple is throwing support behind labeling human-made works across online platforms -- something Instagram head Adam Mosseri believes may be easier to achieve compared to identifying and labeling AI-generated content.

Apple may introduce a photo authentication tool in iOS 27 - Engadget

Reference Image could be a new way to prove provenance in the wake of generative AI. We're looking forward to seeing the new artificial intelligence side of Siri when iOS 27 launches, but it appears that Apple may have additional tools in the works. In its analysis of the latest operating system beta, 9to5Mac noticed some signals that Apple might be working on a photo authentication feature. Apple Reference Image appears to be a tool that will verify that a photo was taken with an iPhone camera. To confirm that a photo is an iPhone capture, select sensor information and metadata for an image can be sent to Private Cloud Compute for verification. A new Reference mode will need to be enabled on an iPhone's camera in order for this authentication option to be available. Resources for checking an image's source in this age of AI slop is more critical than ever. Thus far, most of the tools involve putting a digital watermark onto generative AI creations, such as those in Apple's Image Playground and its other forthcoming image editing capabilities. Google also continues to expand its SynthID AI watermarking program. This approach of Reference Image to create confirmation of content's provenance could be a useful counterpart to those metadata watermarking efforts. Since all this is happening in an ongoing beta test, however, it's possible that Reference Image won't be included when iOS 27 does have its public rollout this fall.

Apple authenticating iPhone photos is just the start

iOS 27 Beta 5 includes references to an upcoming new feature to allow Apple to authenticate iPhone photos. Not many iPhone users will need to use this for reasons we'll get to in a moment, but the move is just the start ... What is Apple Reference Image? AI-generated images and video have rapidly grown in sophistication and realism to the point where it can sometimes be hard to distinguish them from genuine photos. This fact is being exploited by both scammers and those intent on spreading disinformation. An industry body has been set up to fight this, known as the Coalition for Content Provenance and Authenticity (C2PA). This provides a means of identifying genuine photos and enabling viewers to see how they have been modified, whether or not AI tools were used. Over 500 companies have joined the C2PA initiative, while Apple for now appears to be going its own way with the Apple Reference Image (ARI) feature. Anyone who has taken a photo on an iPhone with ARI enabled can upload it to an Apple server for verification that it is genuine. Why is this just the start? Assets found within the latest beta indicate that the ARI feature will be off by default, and that's because most iPhone owners will never need to use it. It's primarily intended for people like journalists and those involved in legal proceedings who may need to be able to produce independent proof that their photos are genuine. But while ARI may have limited usage when it launches, features like this are going to become increasingly common as usage of AI image-generation grows. For example, the EU recently announced a code of practice on the transparency of AI-generated content intended to allow people to easily tell when content is not real. This applies to both images and text (see below). This is currently a voluntary initiative, but one AI giant has already seen which way the winds are blowing and has announced support for it. Anthropic says that all future models of Claude will apply digital watermarking to both text and images. Notably, the company is doing this worldwide, not just within EU countries - a tacit acknowledgement that such requirements are likely to become mandatory globally in the not too distant future. How do you authenticate text? The most eye-opening requirement in the EU initiative is that AI companies must digitally watermark text output as well as images. This can be done using invisible characters, such as the one I just used to replace the space between the words 'invisible' and 'characters.' This would survive copying and pasting, although would be very easy to defeat by using optical character recognition. For example, I have a utility on my Mac that allows me to do as easily as taking a screengrab (installed because of the highly annoying habit some companies have of tweeting statements as images). Some are suggesting that Anthropic may go as far as using particular language patterns in Claude output in order to allow detection even for OCRed text. I could have very much to say about this, but that's beyond the scope of this piece. If this kind of thing does indeed become mandatory, then Apple will need to apply it to tools like Siri AI rewrites. What are your thoughts on this trend? Please share in the comments.

What's up with this 'Apple Reference Image' rumor?

There are plenty of companies adding metadata to AI generated images, and Apple is one of them. In iOS 27, images created with Image Playground will embed the SynthID metadata watermark created by Google. But Apple Reference Image seems to be a tool to help authenticate images another way: rather than just by positively identifying fake images, it authenticates that an image or video is real and has been captured by an iPhone. The feature is not active yet, but when it is, you'll find it in Settings > Camera > Reference Image. Here's how it appears to work: In addition to enabling the feature in Settings, you'll have to take your photos using a new Reference mode in the Camera app (which does not exist yet). Once you do so, a bunch of extra metadata will be saved along with your image. Then, when you want to authenticate it, you can tap a "Reference" badge to send the raw image and all sorts of metadata up to Apple, which uses the Private Cloud Compute servers to determine that the image did in fact come from an iPhone image sensor and has not been edited. It then assigns the image a unique photo ID and sends back that authenticated version. This may sound like a great way to push back against the flood of AI generated and manipulated images online, but it probably won't really make a dent there. Most of that stuff spreads on social media where images are recompressed and often stripped of some metadata before they are shown to other users. Plus, this is entirely opt-in, requires a special mode, and it's unclear which regular image edits can be done (if any) without removing the authentication. Rather, this seems like a way to create a legally robust providence claim for occupations that need to prove a photo is legitimate. Think: journalists, insurance agents, criminologists, and so on. A more robust system is the C2PA (Coalition for Content Provenance and Authenticity), which has already been adopted by standalone camera companies like Nikon, Sony, and Leica along with software companies like Adobe. Google adopted that standard in the Pixel 10 this year. The C2PA standard creates a "nutrition label" of sorts that shows a history of the way an image was captured (or generated) and broad types of edits made along the way. So you can see if an image is fake, or just "edited" to crop it and apply color balance, for example.

Apple may make fake photos harder to pass off as iPhone shots

In this age of AI, fake photos have become a genuine concern. But Apple might be planning on dealing with this with its next major iteration of iOS. Code discovered in iOS 27 beta 5 by 9to5Mac points to an unreleased feature called Apple Reference Image. The system is designed to authenticate that an image originated from an iPhone camera by using information tied to the hardware that captured it. Meaning, iPhone photos could now arrive with a verifiable record of where they came from. Your iPhone camera could vouch for its own photos Reference Image reportedly won't apply to every photo you snap. Users would first need to enable Reference Mode under Settings > Camera > Reference Image, then capture pictures using a dedicated Reference option in the Camera app. So photos taken through the regular camera modes apparently won't contain the provenance information required for authentication later. Recommended Videos When someone chooses to authenticate one of those images, the raw photo and its provenance data would be processed using Apple's Private Cloud Compute infrastructure. Apple would receive specific sensor data and photo metadata, while the system would assign the authenticated image a unique ID. Apple may even be able to revoke previous authentications if it discovers that a particular camera sensor has been compromised. These authenticated pictures could then be shared with other Apple users, whose devices would reportedly be capable of checking their authentication locally. The company is reportedly also working on a Transfer with Provenance feature for moving images over USB while retaining their Reference Image information. Photo authentication is becoming a much bigger deal Apple's alleged new feature arrives as tech companies increasingly search for ways to establish where digital images originated. Google has embraced technologies including SynthID. Meanwhile, standards such as C2PA attach provenance information that can identify how an image was created or modified. But this approach is leveraging the iPhone camera itself. So instead of focusing on identifying AI-generated material, Apple is giving each genuine camera capture a unique identity to establish its origin. Keep in mind that this is still an unconfirmed feature found inside an iOS 27 beta. Apple hasn't announced it publicly, and there's no guarantee it will make the final iOS 27 release this fall. But I'd like to see this one survive the beta.

Apple is working on a way to authenticate that a photo came from an iPhone camera

References included in iOS 27 beta 5 point to a new feature designed to authenticate the camera source of an iPhone photo. Here are the details. Apple working on new Reference Image feature While the challenge posed by synthetic images trying to pass as authentic is not new, it has become an increasingly pressing issue with the rise of generative AI imagery. What used to take a skilled photo editor hours, or even days, can now be accomplished in seconds, with little more than a short prompt based on a half-formed idea. To address this, several companies have already introduced ways to make AI-generated content easier to identify. Google, for instance, developed SynthID, which embeds imperceptible watermarks into AI-generated images and other media. The technology has since expanded beyond Google's own products, with companies including OpenAI adopting it for generated images. Likewise, Meta has introduced labels for AI-generated content across Facebook, Instagram, and Threads, though such systems have faced limitations in reliably identifying manipulated media, even when created by the company's own AI models. Now, it appears that Apple might be gearing up to tackle the problem as well, in addition to "marking up" the metadata of images created via Image Playground. New code in iOS 27 beta 5 shows that the company is working on something called Apple Reference Image, which is designed to authenticate the source of photos using unique data tied to the iPhone camera hardware that captured them. In line with Apple's privacy-focused approach to cloud-based processing of potentially sensitive user data, the company is designing Apple Reference Image in a way that prevents Apple from accessing the raw photo itself. Instead, the system sends the image to Private Cloud Compute for authentication, while only certain sensor information and photo metadata are sent back to Apple. Interestingly, if Apple determines that a sensor may have been compromised, it may revoke prior authentications associated with that sensor. As it stands, the feature is off by default, and once it's live, it will be turned on via , and tapping Reference Mode. Importantly, Reference mode must be used when capturing the photo via a new "Reference" mode in Camera. This means that photos taken normally won't contain the provenance information needed to be authenticated later. Reference mode itself does not appear to authenticate the photo automatically. Instead, users can later choose to authenticate it, at which point the raw image and associated provenance data are sent to Private Cloud Compute for verification. During authentication, Apple assigns the photo a unique ID. Another privacy-focused aspect of the system is that authenticated photos can also be shared with others, whose Apple devices can check locally whether the photo remains authenticated, without telling Apple which images they are viewing. The information included when sharing may vary depending on the method used (AirDrop, Messages, etc.) and whether the photo is an original, edited, or authenticated version. Notably, if the user chooses to share "All Photos Data," Apple says the transfer may also include unique hardware identifiers and any uncropped footage associated with the image. Apple is also adding a "Transfer with Provenance" option for USB transfers. When enabled, authenticated photos transferred to a Mac or PC will include their Reference Image data, while unauthenticated photos may still carry provenance information such as unique hardware identifiers and any uncropped footage. What's your take on Apple Reference Image? Let us know in the comments. Worth checking out on Amazon

Apple tests iPhone photo authentication feature for iOS 27

Apple may add a photo authentication feature called Apple Reference Image to iOS 27, a tool designed to verify that a picture was taken with an iPhone camera. 9to5Mac identified signs of the feature in its analysis of the latest iOS 27 beta, where the tool appeared as part of Apple's work on image provenance. The feature would allow select sensor information and image metadata to be sent to Private Cloud Compute for verification. That process could confirm whether a photo originated from an iPhone camera. A new Reference mode would need to be enabled in the iPhone camera for the authentication option to be available. The feature is being developed as demand grows for tools that can verify the source of images amid the spread of AI-generated content. Current approaches have largely focused on applying digital watermarks to AI-generated images. Apple already uses watermarking-related tools in Image Playground and in other image editing capabilities under development. Google is also expanding its SynthID AI watermarking program. Reference Image could operate alongside those metadata-based watermarking efforts by confirming a photo's provenance through device capture data. The feature remains in beta testing, and it may not be included when iOS 27 is released publicly this fall.

Apple is building a way to prove iPhone photos are genuine as it adds more ways to generate AI slop

Apple Reference Image would send an image captured on your iPhone camera to the cloud for verification. It's a handy tool in the fight against AI slop. Apple appears to be working on a tool that would prove a photo has been captured on an iPhone camera rather than generated via artificial intelligence tools like the company's Digital Playground. A new Apple Reference Image feature has been spotted in the latest iOS 27 beta. 9to5Mac found that the tool uses the metadata attached to the photo that identifies the image and traces it back to the device that captured it. The raw photo can be officially authenticated via Apple's Private Cloud Compute. The image itself doesn't go to Apple as part of the verification process, just the metadata and the sensor information. Users will need to enable Reference mode on their iPhone's camera settings in order to begin authenticating their product. When iOS 27 launches you'll see it at "Settings > Camera > Reference Image > Enable Reference Mode." The report says the feature won't automatically go about its business, but iPhone owners can manually do so on occasions where verifying a photo is necessary - when photographers want to prove that the photo is an authentic capture and not more AI slop. Apple already applies a watermark to images fashioned in its Image Playground app already include a hidden SynthID watermark, so Reference Image will do the same for proper photos. It's not clear how the feature will react to further AI amendments made in the editing process but in an ideal world the verification would be removed. It's also not guaranteed that Apple will roll this tool out in iOS 27 upon its September release, but it may come later in the cycle.