Will DeepFake Replace Everyone?

[Man]

Hello! Today we will talk about what is happening in the world with the advent of deepfakes: how AI set the Eiffel Tower on fire, Taylor Swift went into big cinema, the Pope became a fashionista, and the state decided to legalize…

First, let's look at a few examples of how Deepfake technologies are used.

Paris is on fire​

In January of this year, all social networks were blown up by the news of the burning Eiffel Tower. But it started burning in the summer of 2023, when the 3D effects company Augmented One posted a 9-second video of the burning landmark on its YouTube channel.
In the description, the guys indicated that this was computer graphics. However, Internet users ignored this and began posting the video en masse, passing it off as real.
Then more and more new photos from the "scene" began to appear, and this time the Midjourney neural network did its best. The photos of the burning Eiffel Tower were so realistic that they got millions of views.

Important statement from Oleg Tinkov*​

In the fall of 2021, a fraudulent video appeared online featuring the founder of the bank of the same name from the Tinkоff Bonus page. The businessman in the video promised to give 50% of the investment amount upon registration.
“For example, you invest 20 thousand rubles, and you receive 30 thousand rubles in your work account,” says the fake Tinkov.
The video was created by scammers and distributed on social networks. The main goal was to get money and personal data from gullible users.
*Oleg Tinkov acts as a foreign agent.

Taylor Swift Deepfake Video Scandal​

In January of this year, X, formerly Twitter, was flooded with AI-generated porn featuring singer Taylor Swift. One of the fake posts even racked up 45 million views before it was removed by moderators.
X promised to actively block prohibited content, but to begin with, they simply disabled search for the query “Taylor Swift”.

Fake Pope Party​

The pontiff was first dressed using the same Midjorney neural network last year. The media and users were not bothered by the fact that the photos were posted on the Midjourney subreddit. But even without this, there is plenty of evidence that the photos are fake.
For example, in one of the photographs the Pope has four fingers instead of five, and in another picture the frame of the glasses ends at the level of the ears, without going around them at all, and the cross doesn’t look very believable.
Articles with loud headlines were published by the American New York Post, Forbes and others, as well as the Russian Gazeta.ru, Afisha Daily and RBC. Then users went into a dance when they dressed the bishop in multi-colored outfits and did not limit themselves to a Balenciaga down jacket.

How did the government react to deepfakes and will they replace real people?​

The scandal has reached the American White House. Press Secretary Karine Jean-Pierre issued a statement calling on Congress to draft legislation that would protect people from deepfake porn.
And recently, on February 8, the US Federal Communications Commission banned companies from using AI-generated voices in robot calls. There is no federal law in the US that prohibits deepfakes. But some states have passed laws against deepfake pornography.
But while some deepfakes are being banned, others are being legalized. For example, Universal Music Group and Google are in talks to license artists’ voices and melodies for AI-generated content. The companies are considering developing a deepfake with popular artists’ voices.
Some artists have embraced the development of neural-network-generated music, including singer Grimes, who told Rolling Stone that anyone could use her voice to create songs “without any penalties” — as long as she got half the royalties.

How Should We Live in a World of Deepfakes?​

There are concerns that deepfakes will eventually become so convincing that it will be impossible to distinguish them from reality. Just imagine: scammers making fake videos of government officials, company executives, and even your loved ones. The last one will give you goosebumps.
Ordinary consumers of information and users of services can only hope that with the development of artificial intelligence, not only laws will develop, but also technologies that will limit the spread of fakes. The world is already taking its first steps in this direction.
There are several startups developing technologies to detect fake videos. One of them is Sensity AI from the Netherlands and Sentinel from Estonia. And the popular Intel Corporation has released a product called FakeCatcher, which can detect fake videos with 96% accuracy by observing subtle color changes in the skin. And some companies have committed to inserting digital watermarks into images made with the help of AI.
But don't rely on technological solutions alone. You can take care of your own protection against deepfakes right now.
Check your sources of information. Don't immediately believe sensational news, even if everyone is forwarding it, check the facts with reliable sources.
Be careful on social media. This is the main channel for deepfakes to spread, so do not distribute questionable content, check its veracity before publishing.
Learn to spot deepfakes. Even if you don't have any special tools, you can still spot strange details in photos and videos.
Warn your loved ones. To force a person to provide personal information or transfer money, scammers can call them on behalf of friends or relatives, imitating their voices using special programs. It is worth coming up with a code word for relatives in case someone calls them with your voice.
Try Chad AI — a service where we have collected and adapted the most popular neural networks in the latest versions for users from Russia. Use the service and forget about the inconvenient interface, foreign SIM and VPN.

Source