YouTube has introduced a significant feature that provides creators with enhanced control over how their content is used for artificial intelligence training. Announced on Monday, this new capability allows creators and rights holders to specify which third-party AI companies can utilise their videos to train models. This change comes amid growing concerns from creators about the unconsented use of their material by leading tech companies to develop AI technologies.
Available immediately, the feature can be accessed through the creator dashboard known as YouTube Studio. Eligible creators, defined as those with administrator roles in the YouTube Studio Content Manager, can opt into the new setting and select from a list of 18 pre-approved companies. This list includes notable firms such as Amazon, Adobe, OpenAI, and IBM, all of which are engaged in the development of generative AI models. Creators also have the option to authorise “All third-party companies,” granting broader consent for their content to be used in AI training, regardless of inclusion on the list.
The introduction of this feature follows complaints from creators who reported that their content had been used for AI training by companies like Apple and Google without prior consent or compensation. In response to these grievances, YouTube announced back in September its intention to provide a framework to protect creators' rights concerning their work being employed for AI development.
While this new setting offers a measure of control to creators, YouTube has clarified that Google will continue to train its own AI models on certain YouTube content based on existing agreements with creators. Furthermore, this new arrangement does not alter YouTube's Terms of Service, which prohibits third-party access to creator content without explicit permission, such as through scraping methods.
YouTube aims for this feature to mark a progressive step toward enabling creators to permit AI training on their content while discussing future possibilities of allowing authorised companies access for direct downloads of videos. As part of the rollout, a default setting will ensure that no third-party companies can train on creator videos unless explicitly permitted, addressing creators' concerns regarding possible violations in the past.
Creators around the world will be notified of this new feature through banner alerts within YouTube Studio on both desktop and mobile platforms over the coming days.
In a parallel development, Google’s AI research lab, DeepMind, unveiled a new video-generating AI model named Veo 2, aimed at competing with existing technologies such as OpenAI’s Sora. This release further underscores the intensifying competition within the AI landscape and its implications for content creators and the future of media production.
Source: Noah Wire Services