Deepfake: Why This Technology is Dangerous


Deepfake: Why This Technology is Dangerous

Technology is impacting our lives positively by making it easier. However, not all technologies have positive impact. For example, the use of deepfake is so dangerous and we should be worried about it.

In this article, we explain what deepfake is and what its dangers are. We will also give some pointers on how to detect deepfakes.

What is deepfake or deepfakes?

Deepfake is a word used to describe a video or image in which a person's face is altered, and sometimes the voice is also changed.

Face faking techniques are nothing new, but they have become much smarter in time. Deepfakes have become more believable, thanks to the huge advancement in machine learning.

Changing someone's face in a video is clearly a disaster. This is especially true in the age of the internet as publishing content is as easy as talking to a friend.

The dangers of deepfake

Accurate face faking brings many dangerous deeds to mind, starting from faking political videos to using it in pornographic content.

Take for example this video of the former President of the US, Barack Obama. Although the video does not bring real danger or scandal, but it shows how convincing these deepfakes can be.

The following list brings out some of the dark ways deepfakes can be used and how this technology can become a very dangerous tool for bad people.

Faking adult materials

Deepfake porn is a real thing. Sometimes female celebrities' faces are used in porn videos in a way that can trick some people into believing them. Until now, there hasn't been a real way to stop this from happening.

Apart from celebrities, it actually happens to regular people. Disturbing truth is that it can be done using bots, which can release huge numbers of nude images of different people.

Some apps are working without any guidelines or control over channels that distribute content to millions of people.

Those apps make illegal content distribution much easier when it comes to distributing them to the mainstream, apart from the dark web and dedicated forums.

For example, one Telegram bot was found to distribute over 100,000 images through a Telegram channel, using images that were captured from Facebook accounts and other sources.

Deepfaking is sometimes used to deliberately destroy someone’s life. It’s done by deepfaking an image of a person, most commonly of a woman.

Then the person is blackmailed by threatening to send the deepfaked files to their loved ones or employer.

Hoaxes of high level

Deepfake dangers
The ability to morph change and recomposite them produces some spooky and rather disturbing results. Feels very odd to see a face that's similar to your own, yet looks like a new person...There's almosed a twisted fascination with this, as though it shouldn't be done, but the results are so uncanny that you want to do more. Scary!

There have been internet hoaxes since the web was created. They have become more common during the presence of social media platforms. Hoaxes created using deepfakes are some of the smartest ones.

It’s possible to make public figures say fake messages to trigger something. For example, make people believe in something that is not true or influence a stock price crash. The possibilities are unlimited.

Deepfakes can be more dangerous for less famous people with fewer followers on social media.

In such case, typically there would be more views on circulating fake videos than on the real video on the affected person’s social media account.

With the technology becoming easier and cheaper, deepfakes are becoming smarter as they can deliver the same fake message in various places, styles and tones to be more believable.

Phishing and scamming

As mentioned above, there are ways to replace the faces of people in videos. This means it’s possible to create videos of trusted people to trick others into phishing scams or steal their data or money.

This can have huge impact on individuals or even corporations, who could suffer data breaches or thefts without having any clue about the culprit behind the scam.

Denial of real evidence

Another danger of deepfakes is that the argument can be used to deny crimes. Nothing is easier than saying the video evidence of the crime was a deepfake. Deepfakes can also be used to set up an innocent person.

This needs a solution that allows the authorities at least to determine whether a video is a deepfake or not.

How to detect deepfakes?

How to detect deepfakes

Deepfakes are becoming harder to spot because there is less blurring and occurring artifacts. And the cost of using the deepfake technology is also falling rapidly.

While there is no specific way to determine which video is a deepfake, there are things to notice when you suspect a deepfake video.

The first details you should carefully inspect are the face expressions with the body movement. The deepfake technology is yet to master natural eye movement. The unnatural emotions can hint the video might be a deepfake.

The most reliable way to detect deepfakes is checking the sources. Following trusted sources and verified pages is a great practice in social media.

Any piece of information should be verified whether it comes from a trusted source or a fake video.

In the future, tools like Microsoft's deepfake detection tool can offer additional help, but these tools still need some development.

Coursera WW

Bottom line

Deepfakes are dangerous with numerous potential threats that come with such technology, fueling misinformation problems that are already here.

There are only a few ways to detect them, but hopefully deepfake detection technology will advance soon to a sufficient level.

Read more: Artificial Intelligence AI vs. Machine Learning