Carding Forum
Professional
- Messages
- 2,788
- Reaction score
- 1,176
- Points
- 113
Legal prohibitions and technologies to protect against unauthorized use of images.
Modern generative AI tools have made the creation of deepfake pornography accessible to everyone. A 2023 report from Home Security Heroes, a company that reviews anti-identity theft services, indicates that creating a 60-second deepfake pornographic video requires only one clear image of the face and less than 25 minutes. This is made possible by free tools.
Public attention to this problem was attracted in January 2023 by graphic deepfake images of Taylor Swift, which spread on social networks and gained 47 million views before being deleted. Similar cases affected not only entertainment industry stars like Korean pop stars, but also ordinary people. According to the report, 99% of the victims of such videos are women and girls.
Lee's company, That'smyface, develops image recognition tools for corporate clients so that their logos, uniforms, or products don't appear in pornography. In the future, it is planned to create a tool for scanning the Internet for deepfake images or videos with users ' faces.
Liu is in talks with Meta about a pilot program that will provide automated content moderation. In the future, the tool can become part of the online identification infrastructure, allowing you to check fake profiles in social networks and on dating sites.
49 U.S. states, as well as the District of Columbia, have laws against unauthorized distribution of intimate images, but their effectiveness varies. Laws often require proof of an attacker's intent to cause harm, which is difficult to prove. In different jurisdictions, there is also no consensus on whether the distribution of deepfake pornography is considered a criminal offense or a civil offense, and who should be held responsible – distributors or platforms.
Source
Modern generative AI tools have made the creation of deepfake pornography accessible to everyone. A 2023 report from Home Security Heroes, a company that reviews anti-identity theft services, indicates that creating a 60-second deepfake pornographic video requires only one clear image of the face and less than 25 minutes. This is made possible by free tools.
Public attention to this problem was attracted in January 2023 by graphic deepfake images of Taylor Swift, which spread on social networks and gained 47 million views before being deleted. Similar cases affected not only entertainment industry stars like Korean pop stars, but also ordinary people. According to the report, 99% of the victims of such videos are women and girls.
Reaction and new technologies
This situation causes active actions on the part of women. Nadia Lee, the founder of the startup, said: "If security technologies do not develop at the same speed as AI, we will face a catastrophe." Despite research on the detection of deepfakes, these tools do not keep up with the technologies of their creation. In addition, their effectiveness depends on the platforms willingness to deal with deepfakes, and most of these videos are posted on specialized sites.Lee's company, That'smyface, develops image recognition tools for corporate clients so that their logos, uniforms, or products don't appear in pornography. In the future, it is planned to create a tool for scanning the Internet for deepfake images or videos with users ' faces.
Personal experience and new solutions
The founder of Alecto AI startup, Breeze Liu, became a victim of deepfake pornography in 2020 after discovering more than 800 links to fake videos. The police were powerless, and they had to track and delete the video on their own. Liu decided to use AI to fight AI and created an app to test the use of user facial images on major social platforms.Liu is in talks with Meta about a pilot program that will provide automated content moderation. In the future, the tool can become part of the online identification infrastructure, allowing you to check fake profiles in social networks and on dating sites.
Legislative initiatives
Removing deepfake content from social media is a difficult task, but removing it from pornographic platforms is even more difficult. Sexual assault defense lawyers believe that legislative measures are needed. Susanna Gibson, who founded the nonprofit organization MyOwn after a personal experience with deepfakes, has secured passage of a law in Virginia in 2023 extending current legislation against " revenge pornography." Despite efforts, such laws have not yet been widely adopted at the federal level.49 U.S. states, as well as the District of Columbia, have laws against unauthorized distribution of intimate images, but their effectiveness varies. Laws often require proof of an attacker's intent to cause harm, which is difficult to prove. In different jurisdictions, there is also no consensus on whether the distribution of deepfake pornography is considered a criminal offense or a civil offense, and who should be held responsible – distributors or platforms.
International practice
The United Kingdom passed a law criminalizing the distribution of deepfake pornography in 2023, and the European Union adopted a directive against cyber violence, including deepfakes, with implementation until 2027. Australia and South Korea also have laws against the distribution of such materials. In China, there are restrictions on the distribution of synthetic content, but their effectiveness has not yet been proven.Conclusion
While legislative measures do not produce the desired effect, startups such as Alecto AI and That'smyface come to the rescue. However, these measures are similar to using whistles to protect against attacks-useful, but not enough. Society must take more serious measures to prevent sexual violence in all its forms.Source
