YouTube collaborates with record companies

According to reports, YouTube, a subsidiary of Google, is in talks with major record companies such as Sony, Warner, and Universal to legally use its content to train AI music generators. YouTube plans to provide these companies with cash in an attempt to persuade more artists to allow their music to be used for AI training.

YouTube's AI program

• Testing phase: YouTube started testing a generative AI tool called "Dream Track" last year. It is reported that users can enter text prompts to create short music clips, and Dream Track will mimic the voice and lyrics of well-known singers. However, only 10 singers agreed to participate in the testing phase of Dream Track and were only available for use by a small group of creators.
• New Attempt: YouTube plans to launch a new AI song generator and hopes to sign contracts with dozens of artists. The report points out that new AI tools may become part of the YouTube Shorts platform, competing with TikTok.

Resistance to implementing the new plan

• Artist opposition: Many artists strongly oppose AI music creation, fearing that it will affect the value of their works. Record companies are also concerned that forcing artists to participate could cause controversy.
• Legitimate use: Music company executives have stated that they are working hard to address this issue and will discuss how to legally use copyrights.

Progress of plan implementation

• Negotiation agreement: According to informed sources, the license only applies to selected artists, depending on whether the record company encourages artists to participate in the new program. The amount that YouTube will pay to the record company has not yet been determined, and negotiations are still ongoing.
• Participation Plan: Sony Music is currently in talks with YouTube. Although it has not participated in the first phase of testing, the company hopes to provide some of its music to new AI tools for training, and some artists from Warner and Universal have already participated in the first phase of testing.