On October 7, 2024, Meta announced the development of a new AI video synthesis tool called Movie Gen, which has the potential to create realistic video clips from text descriptions. This technology, while groundbreaking, poses significant risks related to the proliferation of deepfakes, particularly as it could allow users to generate synthetic videos of themselves with just a single photo.
Unlike existing AI tools, Movie Gen can produce longer videos, up to 16 seconds in Full HD resolution, making it suitable for platforms like TikTok and Instagram. The AI leverages Meta's extensive database of images, raising concerns that individuals could be targeted for deepfake videos without their consent.
The implications of this technology are profound, especially given the existing issues with deepfakes in various regions, including South Korea, where students have already been victims of such manipulations. Meta has yet to announce a release date or usage terms for Movie Gen, indicating an awareness of the potential dangers.