Deepfake AI Apps: Pet Editors Hid Sexual Face Swap Features
Quick answer
How can a deepfake app hide inside a pet editor or fitness app?
A harmless-looking app can ask a remote server whether to reveal extra features. Reviewers see only the clean version, so a face swap tool can appear later without a new review. Because a store listing shows little, check whether the photo access an app requests actually fits its stated job.
Picture a free app that promises to turn your dog into a cartoon. Or one that "coaches" your workouts. Deepfake ai apps can hide behind exactly that kind of friendly face, and you would never know from the store page.
Deepfake ai apps can look like pet editors or fitness coaches while hiding a switch that unlocks sexual deepfake tools, so a clean app store listing is not proof your photos are safe.
Here is the short version, texted the way I would text a friend. WIRED reported that hundreds of mundane apps were hiding sexually abusive deepfake features. A deepfake is a fake picture or video made by artificial intelligence (ai, software that learns patterns from huge piles of examples), and it can make a real person look like they did something they never did. If you have photos of your kids, your partner or yourself on your phone, this story is for you.
Deepfake AI Apps Are Hiding Inside Pet Editors and Workout Coaches
According to 9to5Mac's coverage of the WIRED investigation, the network spanned more than 400 connected apps and websites. Many of these deepfake apps were dressed up as pet-photo editors and workout helpers. Together they were downloaded by users more than 350,000 times, and many carried ratings that said they were fine for all ages.
Deepfake apps and the face swap trick: how a harmless screen flips
Here is the part that should bother you. These deepfake apps mostly showed normal content. But a check against a faraway computer (a server, basically the app's remote brain) could switch on the hidden tools. Reporting from IBTimes Australia describes activation tied to behavior, plus location and language. Someone who arrived through a particular ad could see a face swap tool (software that pastes one person's face onto another body) that other app users never saw.
Think about what that means for a reviewer. They open the app, tap around, see a cartoon dog, and approve it. Nothing looks wrong, because on that day nothing is wrong.
Avatar, pet and fitness apps: what the disguise looks like
An avatar (a cartoon or digital stand-in for you) sounds sweet. A fitness coach sounds boring. That is the point. Boring is a great costume. Nobody expects a pet editor to be a deepfake maker (a tool built to put real faces into fake scenes), so nobody checks.
Why Deepfake AI Apps Slip Past Review: The Key Features of the Trick
Our research file puts the core problem bluntly. App stores review a snapshot, but the apps keep changing after approval. This article is part of a series, start with Biometric Face Scan Brazil Ties 280m Phones To Verification.
A security researcher approving an app on day one has zero visibility into what that same app delivers on day 180.
Expert analysis in CaraComp's research file on the WIRED story
Hidden features: what the server check really does
Let me translate. The store sees the app's shell. The shell can ask the server, "Should I show the real menu today?" If the answer is yes, new features appear without any new version going through review. That is the key trick, and it is why a careful look at the listing tells you very little. Good analysis of the app on approval day cannot see what arrives on day 180.
How did the stores do once they learned about it? Apple removed 19 apps from its iOS App Store after WIRED shared details. Google had already removed most of the identified apps before WIRED reached out. And tens of thousands of explicit deepfake ads tied to the network ran on Meta's platforms this year. So the apps were not just sitting in a store waiting. They were being pushed at people.
Why Deepfake AI Apps Matter to Ordinary Families
- Disguise: the app can look like a pet or workout tool, so the listing proves little.
- Delay: some apps were available for about a year before they were pulled.
- Reach: ads pulled users in, so it was not only random store browsing.
- Targeting: according to the Institute for Strategic Dialogue, 99% of sexual deepfakes feature women.
That last point matters. This is not neutral spyware. It is a pipeline that mostly hurts women and girls, and that is why a parent of a teenage daughter should care about deepfake apps that look like homework helpers.
What Makes a Deepfake Maker Different From a Normal Photo Editor?
A normal editor changes how a photo looks: brightness, filters, a funny hat. A deepfake maker changes who appears to be in the picture or video. It takes your face from an uploaded image and builds something new around it, often something sexual, and often without your knowledge. The risky moment is when you upload.
| Question | Ordinary editor | Disguised deepfake maker |
|---|---|---|
| What it does to a photo | Adjusts lighting and color | Swaps faces and creates deepfakes: fake images and videos |
| What the store listing shows | Matches what the app does | Shows a pet or fitness app with ordinary features |
| What users see | The same screens every day | Normal screens for most users, hidden tools for users who meet the server's conditions |
| Why a photo upload is risky | Low, if it stays on your phone | High, if it reaches a server that can switch on abuse tools |
| Who is harmed | Nobody | Someone whose identity is attached to a fake |
Face swap tools and the brand name trap
You may have seen familiar brands pop up when you search for face swap or avatar tools. A well-known name, or a slick site, is not a safety check. The reporting here is about apps that looked unremarkable, and a pet editor with five stars can still hide a switch.
Who is the maker behind the app?
Check the maker (the company or developer named on the store page). Look for a real website, a real address and a history of other apps. A brand new maker with a dozen near-identical apps deserves suspicion. The researchers noted that apps in this network began showing up in late 2025 and rose sharply in spring 2026. That is a factory, not a hobbyist.
What To Do About Deepfake AI Apps Before You Upload Another Photo
I will give you one thing, not ten. If an app asks for access to your photos and the job it claims to do does not need them, stop. A flashlight does not need your photo library. A step counter does not need your faces. A pet editor needs pet pictures, so choose single photos to share rather than opening your whole library. Previously in this series: Facial Recognition In Stores Shoppers Wrongly Tied To Crime.
That check is not perfect. A pet editor legitimately needs a photo, and the hidden features in this story rode along with a reasonable-looking request. So treat the permission prompt as a warning light, not a guarantee. Careful analysis of what an app asks for will catch some of the bad ones, not all.
Image permissions: what to check before you upload
On your phone, open settings and look at which apps can see all your photos. Switch the ones that do not need it to "selected photos" or "none." Then delete apps you have not opened in months. Each one is a possible door.
If you have ever wondered whether a photo or profile is really who it claims to be, that is the exact question face comparison and detection (spotting fakes and checking matches) exist to answer. Careful analysis of a suspicious image, looking at the lighting, the edges of faces and whether an identity holds together, can sometimes suggest whether something was made by software, but it cannot give a guaranteed answer. Detection works after the damage is made. The better move is keeping the original photos from reaching a risky app at all.
Why detection and analysis come second
Here is what most people miss. Tools for detection and analysis of fakes are improving, but they are chasing a very large pile. In 2023, estimates put sexual deepfake videos online at about 500,000. By 2025, that estimate had grown to about 8 million, according to TechXplore. Analysis of a single fake is one thing. Analysis of millions is another. That is why stopping the upload matters more than hoping a detection tool catches it later.
Apple did update its App Review Guidelines on October 16 to name AI-generated nonconsensual intimate imagery and hidden features. That is a good step. But rules on paper do not stop an app that changes after approval.
Deepfake ai apps can pass as pet editors or fitness coaches, so judge an app by whether its photo permission makes sense for its job, not by how harmless it looks. The app, not your photos, should be the thing you doubt.
Recap: Ordinary-looking apps can hide a switch that unlocks sexual deepfake tools, and a clean store page proves nothing. Ask whether the photo access fits the job, and say no when it does not.
Apple removed apps after WIRED raised the alarm, while Google had already removed most of the identified apps before WIRED contacted it. Some had been live for about a year. So between the day an abusive app goes live and the day somebody notices, who exactly was watching your family's photos? Up next: Deepfake Ai Apps Pet Editors Hid Sexual Face Swap Tools.
Deepfake AI Apps: Frequently Asked Questions
Can a deepfake app use multiple face photos of my family?
Yes, if you give it access. Some tools can handle multiple face images at once, which is why a family album is a rich target. If a deepfake app asks for your whole library, that is a red flag. Share only the single photo the app needs, and think twice if the app's purpose does not require faces at all. A pet editor should not care about your relatives.
Can an ai deepfake be built from my voice and audio too?
Technology exists that can clone a voice from a short audio clip, and some deepfake technology combines a fake face in a video with a cloned voice. The WIRED story focused on sexual images and videos, so I will not claim more than that. Still, the same habit applies: do not hand audio, photos or videos to an app whose job does not need them.
Is an ai avatar generator the same as a deepfake generator?
Not always. An ai avatar is a stylized digital version of you, and many are harmless. A deepfake generator puts a real person's face into fake content. The line blurs when an avatar app also offers a face swap tool. Read what the app lets you create, and be careful if it can produce realistic media of real people.
Does a creative photo editor count as a deepfake app?
Usually no. A creative editor that fixes lighting, crops or adds filters is doing ordinary edits. It becomes a deepfake app if it can swap faces or generate new videos of real people. The trouble in this story is that the labels did not match the features, so watch what the app asks to access, not just what it advertises.
Does artificial intelligence make deepfake detection impossible?
No, but it is a race. Detection tools look at things like skin texture, odd lighting and the edges of faces, and they can catch many fakes. They struggle at scale, and media spreads fast. Artificial intelligence (ai) helps both sides. That is why prevention, meaning not uploading to risky apps, matters more than hoping a fake gets caught afterward.
Does gender matter in who is targeted by deepfake applications?
Yes. According to the Institute for Strategic Dialogue, 99% of sexual deepfakes feature women. This is not random, which is why deepfake apps that hide abusive tools are a gender-based harm. Parents of girls, and women generally, have extra reason to guard photos and to check which apps can see them.
How do I know if a deepfake app has my photo?
Honestly, you often cannot tell. Once a photo is uploaded to a server, you do not control what happens to it. Check which apps have photo access on your phone, remove the ones you do not need, and avoid sharing family photos with unfamiliar apps. If you find a fake of yourself, save the evidence and report it to the platform.
Ready for forensic-grade facial comparison?
Full forensic reports with detailed similarity scoring. Results in seconds.
Run My First SearchMore News
Facial recognition in stores: Shoppers wrongly tied to crime
A computer can mistake you for someone else, and a store can treat that guess as fact. Here is what to say and ask for if it happens to you.
biometricsBiometric Face Scan: TSA Lets Fliers Say No at 250+ Lanes
Airports are racing toward face-based travel, but your face isn't always your passport. Here's what to check before you agree to a scan.
privacyIdentity Verification Process: 4 Head Turns, Zero Proof
Apple's new selfie age check for iPhones sounds privacy-friendly. But the question families should ask is who can prove the photo is really deleted.