YouTube is developing tools to detect AI-generated likenesses of actors, musicians, and other artists, Amjad Hanif, the company’s vice president of creator products, said in a blog post this week. Hanif also reiterated that YouTube is “committed” to developing AI tools responsibly.
The first of the new tools is “synthetic singing” identification within Content ID, a system for identifying copyrighted music that was launched in 2007. Mashable reported in 2016 that YouTube had paid out billions of dollars to rights holders through Content ID, and Hanif reiterated that in the post.
The new technology will apparently use AI to generate and detect voices that resemble musicians, and a pilot program is expected to begin next year.
Mashable Lightspeed
YouTube expands restrictions on weight-training and fitness videos for teenagers in Europe and the UK
YouTube is also developing technology to detect AI-generated content that includes the faces of actors, creators, athletes and others.
In addition to the announcement, Hanif also noted that YouTube and Google (YouTube’s parent company) use uploaded content to “improve your content experience” through the development of machine learning and AI tools. The company is “committed” to doing this “responsibly,” but did not provide details. The post also said that YouTube has safeguards in place for the use of YouTube’s own generative AI tools, like Dream Screen, including blocking certain prompts that violate policies (though it did not specify which prompts).
YouTube is also (re)committed to ensuring that third-party apps comply with YouTube’s terms of service, which prohibit certain uses of creators’ content, such as scraping.
“We are committed to working with our partners to ensure that future progress amplifies their voices, and will continue to develop guardrails to address concerns and achieve our shared goals,” Hanif wrote.
topic
Artificial Intelligence YouTube
