YouTube AI opens its deepfake video detection tool to everyone

YouTube is opening up the facial similarity detection tool it developed against fake videos created with artificial intelligence to a wider user base. According to the company’s statement, in the coming weeks, all content creators aged 18 and over will be able to check whether their images are used without permission in artificial intelligence videos. The platform states that thanks to this system, users can more easily initiate action to remove unauthorized content containing their own images.

In the announcement published on the YouTube community page, the company stated that the goal is for users to “have a greater sense of trust.” Although the tool technically appears to be aimed at content creators, YouTube spokesperson Jack Malon told The Verge that the system has a wider range of uses in practice. Accordingly, users who have been producing content for a long time and people who have newly joined the platform will be able to benefit from the same level of protection. Thus, not only major content producers but also everyday users will be able to intervene if their faces are used in misleading or malicious videos.

This expansion, which comes at a time when it is becoming increasingly difficult to distinguish videos created with artificial intelligence from real images, attracts attention. Copying the images of celebrities, politicians and ordinary users without permission and using them in fake content, especially on social media platforms, has come to the fore frequently lately. In addition, there are examples where some brands or advertisers can use content producers’ images in promotional videos without permission. YouTube’s new system aims to detect such situations faster.

YouTube introduced the tool in question for the first time in 2024. The system later began to be used with a limited user group towards the end of 2025. At first, only YouTube Partner Program members could access it. In order to participate in this program, content producers had to reach a certain number of subscribers and meet viewing requirements. The platform later made the tool available to users with public profiles such as journalists and politicians.

Recording will be done via YouTube Studio

Users who want to access the new system must log in to YouTube Studio on their computer. Users can start the registration process here via the “Similarity” tab in the “Content detection” section. In the continuation of the process, it is required to scan the QR code from the phone, upload an official ID and verify with a selfie video. After these verification processes are completed, YouTube starts scanning the videos uploaded to the platform to see if they match the user’s face.

Detected possible matches are listed under the same panel. Users can review the relevant videos here and request removal if they deem it necessary. Additional information regarding the context in which the image is used can also be shared during the application. In addition, YouTube also asks for an evaluation as to whether the user’s voice is imitated in the video. However, the current system cannot make automatic detection based on sound alone. Decisions regarding voice imitations are further reviewed along with user feedback.

This step taken by YouTube shows that the scope of security measures developed by platforms against artificial intelligence-derived content has expanded. Digital identity protection has become one of the priority topics of social media companies, especially as deepfake technology has become more accessible. Despite this, issues such as the possibility of mismatching such tools, the speed of review processes and the finalization of removal requests are among the main factors that will determine the impact of the application.