Several AI apps are out there to make the life of people easy, but, there are other AI Bot Apps too to fulfill irrelevant and immoral purposes. This article is about AI Bot deepfake app.
This deepfake bot uses artificial intelligence to undress the photo of women and young girls on the messaging app, Telegram. Thus, deepfake utilizes machine learning and AI to create realistic looking images which are false.
What does the AI Bot app actually do?
The deepfake bot was launched on July 11, 2019, So, this is very similar to DeepNude also introduced in 2019, it allows you to upload a picture of women and get you her back her photo seemingly nude. The algorithm behind deepfakes, to swap the women’s clothes for highly realistic nude bodies is generative adversarial networks. The lesser clothed the victim, the better. The bot does not work on men.
The DeepNude provided you the service for $50 but deepfake provides you the service for free. Sometimes you need to pay 100 rubbles (approximately $1.5). This is for other benefits such as removing the watermark on the ‘stripped’ photos or skipping the processing queue.
Cybersecurity specialist, Sensity AI investigated this case. They found that by July 2020, the bot targeted and “strip” at least 100,000 women. The majority of whom likely had no idea. “Usually the targets are young girls,” says Giorgio Patrini, the CEO and chief scientist of Sensity, who coauthored the report. “Unfortunately, sometimes it’s also quite obvious that some of these people are underage.”
Around 70 percent of the targets are private individuals. The users took their photos from social media accounts or private material. Some of the targets appeared to be underage as already mentioned.
Survey regarding the users
The deepfake bot is linked to seven Telegram channels with a combined total of over 100,000 members. (This number doesn’t account for duplicate membership across channels, but the main group has more than 45,000 unique members alone.)
A poll of 7,200 users showed that roughly 70% of them are from Russia or other Russian-speaking countries. The victims, however, seem to come from a broader range of countries, including Argentina, Italy, Russia, and the US.
The majority of them are private individuals, whom the bot’s users say they know in real life or whom they found on Instagram. The researchers were able to identify only a small handful of women. They tried to contact them to understand their experiences. None of the women responded according to Patrini.
The researchers also reached out to Telegram and to relevant law enforcement agencies, including the FBI. Telegram did not respond to either their note or MIT Technology Review’s follow-up request for comment. Patrini says they also haven’t seen “any tangible effect on these communities” since contacting the authorities.
AI Bot App Deepfake revenge porn
Abusers have been using pornographic imagery to harass women for some time. In 2019, a study from the American Psychological Association states that one in 12 women end up being victims of revenge porn at some point in their life.
A study from the Australian government, looking at Australia, the UK, and New Zealand, finds the ratio is as high as one in three. Deepfake revenge porn adds a whole new dimension to the harassment, because the victims don’t realize such images exist.
Celebrities and individuals of very high-profile have been victimized in different cases by deepfakes. The technology first grew popular in the deep recesses of the internet as a way to face-swap celebrities into porn videos, and it have been used as part of harassment campaigns to silence female journalists.
Patrini says he’s spoken with influencers and YouTubers, as well, who have had deepfaked pornographic images of them sent directly to their sponsors, costing them immense emotional and financial strain.