Cybersecurity expert provides tips on how users can protect themselves from becoming victims of deepfake scams

First introduced to the public in 2016, the advanced photo and video manipulation technology called deepfakes has advanced over the last few years so much that now anyone can face-swap themselves into movies or music videos to create incredibly realistic clips. All you need is a dedicated app and a selfie.

However, it’s all fun and games until some stranger sends you a forged nude of you and threatens to put it on social media for your friends and family to see. That’s exactly what’s recently been happening on the Telegram messaging app, where thousands of women were victimized by having the clothed parts of their bodies in their regular photos replaced with nudity and then posted online.

“The case shows how far deepfake technology has come. On the one hand, it’s affordable to anyone and easy to use. On the other hand, it’s violating and invading not only celebrities’ but private people’s lives too,” warns Oliver Noble, a cybersecurity expert at NordLocker, an encryption-powered data protection solution.

Technologies behind the creation of deepfakes

Even though computer manipulation of visual media is nothing new, the technology is improving rapidly. Before apps and bots generating deepfake content became widely accessible for download, the technology required many minutes of the source video, hundreds of frames of target images, and long hours of computing power. Now, the technology can generate a realistic-looking fake video in a couple of seconds from a single photo captured by any color camera.

Deepfakes use machine learning technology that applies neural net simulation to massive data sets to create fake images. Artificial intelligence effectively detects the source face and its facial features at different angles to place it onto the target as if it was a mask. Some techniques only change the visual features of the face, while others change the facial movements of the target to make better forgeries.

Why are deepfakes scary?

Computer-generated media could be used in socially beneficial ways, such as to significantly reduce costs and time on producing artificial videos like commercials or to help students and artists develop innovative forms of creative expression. Unfortunately, it could also be used to cause harm, and that’s what it’s used for most often.

There are some reasonable concerns that deepfakes can disrupt elections, cause confusion and a divide between people, shift public opinion, and undermine confidence in the news of major current events. A great example of this was a deepfake of Barack Obama that surfaced in 2018 on YouTube, in which the president was calling Donald Trump names. The video was produced to show the dangerous consequences of deepfakes and how they are used to disseminate fake news by using famous people’s faces.

Another scary instance of deepfakes is cybercriminals exploiting voice cloning techniques in social engineering scams. Last year, fraudsters used an audio deepfake to convince a CEO to transfer $243,000 to a scammer’s bank account.

But perhaps the most unpleasant situation deepfakes can put you into is non-consensual sexual images. “Fake nudes are deployed to harass, threaten, humiliate, and extort victims,” explains Oliver Noble. “Users of certain deepfake apps can take your photo and “strip” clothes off you to fabricate pornography. Thus, you need to be extra careful when uploading your photos to social media or storing them unprotected on your phone and computer.”

5 easy steps to protect your photos

The rising issue with deepfake technology once again proves how important it is to keep your private data to yourself and share it only with people you trust. Luckily, protecting yourself against falling victim to non-consensual sexual images doesn’t require any technological knowledge. Oliver Noble shares his five easy tips to reduce your vulnerability.

In order to avoid being compromised in deepfake pornography, don’t post photos of your full-sized self. Upload a nice portrait instead so that apps or bots have nothing to remove clothes from.
Set your social media settings to private so that your personal information and photos are visible only to your friends and family instead of everyone on the internet.
Don’t connect with strangers online. Before befriending someone you don’t know on social media, go through the mutual friends’ list or things you have in common.
Protect your pictures with encryption, which gives you a choice over what you share and what stays private. Better yet, store all your photo albums in an encrypted private cloud such as NordLocker for automatic sync, so in case your device gets hacked or lost, no one can access your precious snaps, and you can easily get them back.
Stay alert when communicating online. Don’t get tempted to send your images to strangers and never click on obscure links they send. Keeping off from uploading your pictures to shady websites should be a no-brainer too.
Deepfakes are becoming mainstream, and it’ll soon be impossible to tell if a video or audio recording is authentic or fake. Luckily, deepfake detection solutions have been advancing, and some governments have even been trying to curb this phenomenon by passing special laws to deal with falsehoods and manipulation online that affect public interest.