Google Leverages YouTube Videos for Veo 3 Training, Creators Cannot Opt Out

Google is leveraging YouTube content to enhance the capabilities of its latest AI model, Veo 3. According to CNBC, content creators do not have the option to exclude their videos from being utilized in this training process.

Launched in May 2025, Veo 3 has quickly gained prominence across social media platforms due to its increasingly realistic outputs. This progression has sparked curiosity regarding the methods employed by Google to achieve such advancements.

As detailed in a report by CNBC, the source of this data primarily stems from YouTube videos which serve as training materials for the AI model.

No Opt-Out for Creators

Not only is YouTube content integral to training the Veo 3 audio and video generator, but it also plays a role in enhancing Google’s ChatGPT alternative, known as Gemini. Google has clarified these details to CNBC, stating that while YouTube videos are utilized, not every upload on the platform is subject to use.

“We’ve always used YouTube content to make our products better, and this hasn’t changed with the advent of AI, ” said a spokesperson for the company.

Furthermore, the spokesperson emphasized the company’s commitment to establishing stringent safeguards to protect content creators’ identities and likenesses in this evolving digital landscape.

Google

Back in September, YouTube announced that uploaded content might be utilized to “improve the product experience… including through machine learning and AI applications.” Unfortunately, creators currently lack the ability to disable this feature.

In December, the company rolled out a significant update concerning how videos can be employed for AI training by third-party models. Through the updated dashboard, creators can opt to permit companies to scrape their videos and select which specific videos may be shared. Notably absent from this selection is Google’s Gemini model, as it is a proprietary system.

Source & Images