YouTube extends AI deepfake detection to all adult users
YouTube has announced a major expansion of its AI-generated deepfake detection system, making it available to all adult users. Until now, the tool was reserved for prominent creators, journalists, politicians and public figures, but it is now accessible to anyone who wants to protect their digital identity. The aim is to curb the spread of manipulated videos that use someone’s face without their consent.
The mechanism is based on a verification process that involves a video selfie and biometric facial recognition. Once the feature is activated, available in YouTube Studio in the dedicated “content detection” section, the platform starts continuously scanning videos uploaded to YouTube. When the algorithm finds a match between the user’s face and suspicious content, it sends an automatic notification, allowing the person concerned to step in.
When a possible unauthorised use of someone’s image is detected, the user can review the flagged content and decide whether to submit a takedown request. Each case is assessed individually, taking into account factors such as how realistic the video is, whether there are labels indicating the use of AI, and the context in which the face appears. Satirical or parody content is still excluded from automatic blocking and continues to fall under the exceptions provided for in the platform’s guidelines.
For the time being, the system focuses exclusively on facial recognition and does not cover voice or other biometric traits. This is one of the main limitations of the technology, especially as many deepfakes combine image and audio to appear more convincing. Access to the service does not depend on subscriber numbers or a channel’s popularity, only on being over 18-years-old and opting in.
YouTube has made it clear that users can disable the service at any time and request the deletion of their biometric data. However, rolling the system out at scale raises fresh questions about privacy, recognition accuracy and error handling, particularly for less well-known faces. The challenge will be to strike a balance between protecting people’s identities and the risk of false positives or overly rigid automated systems.
GreenMe
Subscribers 0
Fans 0
Followers 0
Followers