Within the framework of its new guidelines for synthetic content, YouTube will go on the hunt for deepfakes and songs generated with artificial intelligence. The platform announced the implementation of new measures in order to attack a growing problem that is causing great concern in the entertainment industry.
According to the company explainedin the coming months will begin accepting applications for remove videos that include AI-generated material that “simulates identifiable individuals”. In this sense, YouTube will place special emphasis on deepfakes that replicate the face or voice of certain people.
The video portal explained that it will analyze different factors before making a decision, and that not all of the synthetic content in question will be removed. Initially, YouTube will focus on deepfakes involving “public officials or well-known individuals.” That is, politicians or actors, singers, influencers and artists of all kinds.
One of the elements that will be taken into account when analyzing this issue will be whether the material meets the purposes of satire and parodyor if they seek to use the appearance of public figures to misinform or promote scams.
As for songs generated with AI, YouTube will pay attention to those that “imitate an artist’s unique voice when singing or rapping”. Record companies, distribution labels and singers’ representation agencies may request the removal of this content.
YouTube introduces new tools to prevent misuse of AI
As with the deepfakes, not all material linked to AI-created music will be removed from the platform automatically. The company belonging to Google will allow the use of these creations if they are part of informative content or that analyze the use of digitally created voices.
Initially, the tool will be available to record labels and distributors who are collaborating with YouTube in its first musical experiments with artificial intelligence. But the idea is to expand access to more companies as time goes by.
Music created with artificial intelligence is a hot topic of debate these days, not just on YouTube. Let us remember that Universal Music Group asked Spotify and Apple to prevent the extraction of lyrics and melodies from their artists’ songs. According to the corporation, this information was then used to train artificial intelligence models that could generate songs imitating the performers or composers without paying for copyrights.
In May, Spotify acknowledged removing tens of thousands of AI-generated songs. But that was not the only problem, but also the existence of bots that “inflated” the reproduction count to earn more money. Likewise, Google has experimented with a tool called MusicLM, intended to create music from a text description. However, he decided not to release it to the public due to the high probability that the results were replicas of materials protected by copyright.
For now, YouTube begins a new stage seeking a more responsible use of artificial intelligence tools among users. Additionally, offer the option to request the deletion of deepfakes and other possibly illegal synthetic content, it will surely be useful to those in Mountain View to cover yourself in case someone wants to take legal action against you.