Do you know what deepfake is? It’s a technique that emerged in 2017 that allows the content of an image or video to be hijacked, very often for malicious purposes.
What is deepfake?
The term deepfake is a crossword formed from deep and false learning. It is understood that he is referring to a deeply credible false content.
This phenomenon is said to stem from a technique invented in 2014, in Montreal, by researcher Ian Goodfellow. Using GAN (Generative Adversarial Networks) technology, two algorithms confront each other: one aims to create counterfeits as reliable as possible (called samples), while the other attempts to detect the real from the fake. The two algorithms work together, and as the number of samples available increases, so does the accuracy.
While the creation of audio deepfakes requires considerable hardware resources and is therefore mainly used by professionals, artificial intelligence (AI) now enables the general public to easily create deepfakes without the need for special technical knowledge. Thanks to apps like FakeApp, whose algorithm analyzes facial movements, we can create images or videos that make us look younger or older. Technological developments are making algorithms more realistic than ever, enabling us to create more and more compromising and problematic fake content… and more and more.
The new threat
With all the fake news sweeping social networks and even traditional media, the growing presence of deepfakes brings even greater dangers: manipulation, disinformation, humiliation, defamation… and threats to national security! In April 2018, Jordan Peele, director of the film Get Out, and Jonah Peretti, CEO of Buzzfeed, demonstrated the dangers of deepfakes with their video of Barack Obama giving a fake speech.
Mostly known on social media as a means of damaging international personalities by producing false images and videos showing them in embarrassing situations, think of Daisy Ridley, Gal Gadot, Katy Perry and Taylor Swift who were victims of pornographic deepfakes or Mauricio Macri and Angela Merkel who had their faces replaced by those of Adolf Hitler and Donald Trump respectively, deepfakes can be found on pornographic sites, obviously used without people’s knowledge, and even more worrying is a number of them use images of minors.
Deepfake Porno
On December 4, 2020, The New York Times published The Children of Pornhub, a bombshell article by journalist Nicholas Kristof that described how the porn content platform monetized underage rape, revenge porn and hidden cameras in women’s locker rooms, to name just a few. What all these videos have in common is the absence of consent. Although these videos showed real people, deepfake porn is of growing concern to experts due to the accessibility and ease of use of the technology to create these fake videos. This threat is anything but benign, as it mainly facilitates violence and harassment against women and girls.
Geographic Deepfakes
In April 2021, US technology, news and media site The Verge warned governments about geographic hypertrategy, claiming that geopolitical adversaries could use false satellite imagery to deceive their enemies. An analyst at the National Geospatial-Intelligence Agency has envisioned a scenario in which military planning software is fooled by false data showing a bridge in the wrong place. The Verge also argued that deepfake could go so far as to discredit testimonies based on real satellite images. Citing the Uyghur concentration camps in China, he explained that the Chinese government might try to protect itself by claiming that satellite images of these camps are in fact faked thanks to deepfake; a real danger to national security.
How to protect yourself
In September 2019, in response to the scale of the situation, Google published a database – constantly updated – of over 3,000 deepfake videos to enable engineers and researchers to create automated AI-based detection tools. For its part, Facebook’s FAIR lab is trying to develop “depersonalization”, an AI that would apply a filter to videos to prevent them from being altered by facial recognition software. Accompanied by Microsoft and associated with renowned institutions (MIT, Oxford University, University of Maryland, UC Berkeley), the social networking giant has launched the Deepfake Detection Challenge, a competition aimed at creating software capable of discovering faked videos. For their part, researchers at the University of Washington suggest demystifying the absolute reliability of satellite images and raising public awareness of the potential influence of geographic deepfake to limit the impact of this technology. Finally, according to Solange Ghernaouti, a specialist in cybersecurity and cyberdefense, the solution would be to “digitally tattoo” digital images and videos to validate their content. Software such as InVid and Amnesty Youtube Dataviewer enable journalists to determine whether a video is faked or manipulated.
In conclusion, it’s also up to the various legislators to put in place laws banning content that doesn’t mention the use of video and audio overlays, as China has been doing since January 1, 2020.
Don’t believe everything you see and read on the Internet.
Photo by: Joshua Rawson-Harris
