Lawyer Francesca Gollin, who heads the legal department, together with the following legal assistants Ana Gheorghita, Ashny Young, Olga Ruiz Pilato, and Xinyu Chen, have produced the first report that will serve as an official guide for combating the deepfake phenomenon worldwide. “The issue is more relevant than ever,” explains lawyer Francesca Gollin, who led the research team that produced the report, “the scale of the phenomenon is swelling with dizzying numbers: as of June 2019, the number of deep fake videos detected by IBM was only 3,000. In January 2020, that number had increased to 100,000. As of March 2020, there are over one million deepfake videos on the internet. According to a study by Deeptrace, in December 2018, there were 15,000 deepfake videos. At the Cyber Rights Organization, we decided to prepare protection guidelines for survivors and victims of deepfake and a toolkit for citizens who want to understand and protect themselves from deepfake”. These figures do not seem to smooth out as the years go by: indeed, the number of deepfake videos reached 558,000 in June 2019 and exceeded one million in February 2020. The trend continues upward today.
Director Annachiara Sarto comments on the release of the report: “We are proud to publish this report on the deepfake phenomenon. A text that will be able to help first of all the victims, but also journalists, and professionals on the field to better understand the phenomenon and also to raise awareness about it in our lives and in our societies”.
The deepfake in detail: a toolkit for prevention and protection
Over the past 20 years, the use of deepfakes in multimedia has increased significantly and has had a direct impact on people’s lives in a variety of ways. “Looking back, the first traces of deepfakes,” explains lawyer Francesca Gollin, “were observed in 1997, when Christoph Bregler, Michele Covell and Malcolm Slaney developed the Video Rewrite program, which aimed to edit existing footage and create new content starring a person speaking words not uttered in the original video.”
Exactly 20 years later, deepfakes gained maximum popularity when an anonymous user on Reddit, a popular Twitter-like social network, published AI software that used existing algorithms to create highly convincing fake videos that could fool anyone. It was 2017 when this unknown user contextually decided to make this algorithm available as free software for public use. As a result, many more users and companies started working on developing applications like FakeApp, which simplified the process and turned deepfake into a mass phenomenon. Today, the development and implementation of deepfakes on the Internet is dominated by machine learning and artificial intelligence. For this reason, CRO decided to take an authoritative position on the topic by analyzing every aspect of the phenomenon.
Cyber Rights Organization report: Social media, pornography and reputation
“The organization’s legal team,” explains Annachiara Sarto, “decided to focus on some aspects that are very important to our mission. Namely, the way in which deepfake has characterized the following areas over the years: social media, pornography, legality and the individual’s web reputation.”
In fact, the CRO’s report opens with a careful examination of the impact of deepfake on the landscape related to the spread of fake news on social media for propaganda purposes. As a recent Unesco study shows, very often videos of well-known people made in deepfake and gone viral on social media have contributed to suddenly catalyzing user consensus on highly polarized issues by increasing the overall level of disinformation.
“On the other hand, regarding the pornographic aspect of deepfake,” explains lawyer Francesca Gollin, “our focus was on the use of this software for the delegitimization of women’s dignity and their subsequent sexual humiliation.” Indeed, the CRO Cyber Rights Organization report states, “Sex has often been used in the context of deepfake technology to create massive amounts of sexual video produced without the consent of the persons portrayed’ produced by misogynistic users. In this sense, the universe of digital platforms has made millions of dollars from the viewing, sharing, and streaming of deepfaked images that sexualize women.”
Finally, it is interesting to note how the use of deepfakes can also have an impact on the image damage aspect for individuals or companies. The cyber rights organization’s report cites as an example the case of a deepfake against Belgian Prime Minister Sophie Wilmes carried out by Extinction Rebellion activists, in which the fake prime minister linked Covid-19 to the climate crisis in a video. “These defamatory videos implemented against political figures, companies or ordinary citizens,” explains Andrea Baggio, EMEA CEO of ReputationUP Group and co-founder of CRO, “can have a very serious impact on people’s lives and activities, sometimes irreparably violating their dignity.”
The position statement of Cyber Rights Organization
“We always condemn deepfakes and any other form of altering reality that can violate the dignity of the individual,” concludes Director Annachiara Sarto, “with our direct Helpline, which is open to everyone, we are working to digitally protect the privacy of deepfake victims by providing legal and technical assistance at the time when survivors need it most.” Lawyer Francesca Gollin concludes: “We respond to deepfake cases promptly through investigation, research, monitoring, and case-by-case assessment. We believe that the Cyber Rights Organization’s response to deepfakes contributes to a safer and more secure online world for people.”