2 min read
Artificial intelligence (AI) has significantly contributed to the surge in deepfake content, raising concerns about misinformation. In response, Meta has introduced Video Seal, a tool designed to label AI-generated videos across its social media platforms, ensuring greater transparency and helping users distinguish authentic content from AI-created media.
Sumsub, an ID verification tool, says that from 2023 to 2024, the number of deepfakes globally increased by four times. Fraudulent actions on the internet, like identity theft, account takeovers, and sophisticated social engineering campaigns, are made up of 7% of these deepfakes.
This watermarking tool is an expansion of Meta’s other successful offerings, such as Audio Seal and Watermark Anything.
“We developed Video Seal to provide a more effective video watermarking solution, particularly for detecting AI-generated videos and protecting originality,” TechCrunch quoted Pierre Fernandez, AI research scientist at Meta, as saying.
Video Seal is related to previous technologies introduced by Microsoft and DeepMind, such as SynthID, which is used to categorize videos created by artificial intelligence.
“While other watermarking tools exist, they don’t offer sufficient robustness to video compression, which is very prevalent when sharing content through social platforms; weren’t efficient enough to run at scale; weren’t open or reproducible; or were derived from image watermarking, which is suboptimal for videos,” explained Fernandez, highlighting the advantages of Video Seal.
Video Seal stands out because it can insert a secret message into videos that may be deciphered at a later time, allowing users to trace their origins.
