Priya Patel
TLDR: Deepfake Porn Is a Huge Problem; These Tools Could Help Stop It
- The number of AI-generated deepfakes is increasing rapidly, with most being pornographic.
- Tools such as digital watermarks, defensive methods, and regulations are being developed to combat the issue.
Deepfake videos, images, and audio generated by AI have become a major problem, particularly in the form of pornographic content. The technology behind deepfakes has made it easier for individuals to create fake content with little technical knowledge. A study found that 96% of deepfake videos were pornographic, and the volume of such content has greatly increased. This has led to real-world consequences, including a British teenager taking her life after deepfake pornographic images of her were circulated.
To combat this issue, various tools and methods have been developed. Digital watermarks have been endorsed as a solution to label AI-generated content, making it easier to identify and remove harmful material. Platforms like Google and Meta are implementing digital credentials to clarify the origins of content. OpenAI plans to include visual watermarks and metadata in line with industry standards to provide more transparency.
Defensive tools like Nightshade, created by researchers at the University of Chicago, add imperceptible signals to images that make them unusable when processed by AI. Regulations are also being implemented to protect victims of deepfakes, with some states enacting legal protections, and the introduction of bills like the DEFIANCE Act that allow victims to take civil action against creators and distributors of sexual deepfakes.
While legislation and tools are important in addressing the issue, preventing deepfake porn entirely may not be possible. The focus is on introducing friction to deter individuals from creating and sharing harmful content intentionally. Governments, search engine providers, AI tool developers, and social media platforms all have a role to play in combating the spread of deepfakes.
