YouTube rolled out a slew of generative artificial intelligence products aimed at a broad audience, an effort from the video platform to show its massive investments in AI are bearing fruit.
The features, announced Tuesday, go well beyond the standard AI editing tools that YouTube and its social media rivals have introduced in recent years. Many of the latest tools are tailored specifically to creators and shortform segments known as YouTube Shorts, a sign that YouTube is taking ever-bigger swings to compete with ByteDance Ltd.’s TikTok and Instagram owner Meta Platforms Inc.
Several of its new offerings are powered by Veo 3 Fast, a Google DeepMind model that can in seconds generate realistic video and audio, including dialogue and sound effects, from a simple text prompt on a phone.
“Twenty years ago, YouTube launched with the idea that everyone should have the opportunity to create and find a global stage,” Johanna Voolich, YouTube’s chief product officer, said in a blog post Tuesday timed to the company’s Made on YouTube event in New York. “Since then, we’ve seen creators shape culture and entertainment in ways we never thought possible.” Voolich added that YouTube has now paid out more than $100 billion to creators globally over the last four years.
Creators will soon be able to use Veo 3 Fast to add backgrounds, props and special effects to YouTube Shorts. Starting early next year, podcasters will also be able to use Veo to easily generate video clips related to their audio. And those hosting video podcasts — which YouTubers are now watching for 100 million hours each day as the platform far outpaces traditional TV — will in the coming months be able to use AI to create highlight reels of their shows to share widely across social media.
Beyond Veo, YouTube’s conversational AI tool, Ask Studio, will soon give U.S. creators feedback and analytics on their content, becoming their “ultimate creative partner” and “a trusted companion that Creators turn to first,” Amjad Hanif, YouTube’s vice president of product management for creator products, said in a blog post.
YouTube is also using AI to test ways to better sync its dubbing technology, which the company says has translated more than 60 million videos into 20 languages to help creators reach wider audiences. Soon, a “speech-to-song” capability powered by DeepMind’s AI music model Lyria 2 will let creators turn words or phrases from a video into music to accompany it.
Brands stand to benefit from YouTube’s generative AI push as well: The platform will soon use AI in advertisers’ Google Ads dashboard to recommend creators for brands to work with. It will also use artificial intelligence in YouTube Shopping to make it easier for sellers and creators to tag products in their videos.
YouTube parent Alphabet Inc. is not alone in more deeply integrating generative AI into its social media strategy and investing more broadly in AI talent. Meta, which recently went on a monthslong hiring spree to build out its new superintelligence lab focused on artificial intelligence, is leaning into AI-generated advertisements on Instagram and Facebook. Elon Musk, who recently merged his social network X with his AI startup xAI, similarly plans to use AI to overhaul X’s ad business and fact-check posts on the platform.
The AI gold rush is creating fresh challenges for users across every social media platform, from unchecked misinformation to an explosion of video and audio using people’s faces or voices without their permission. As one safeguard aimed at addressing that issue, YouTube is soon expanding creators’ access to a detection tool that will scour the site for AI-generated videos misusing creators’ likenesses and make it easier for that content to be removed.
“Our goal is to build AI technology that empowers human creativity responsibly, and that includes protecting creators and their businesses,” Hanif said in the blog post.