American country singers, Romance novelist, Video game artists and voice actors are calling on the US government to address the threat that artificial intelligence poses to their livelihoods.

Wrote one podcaster who fears his voice is being copied by AI in a recent letter to the U. S. Copyright Office: "Please regulate artificial intelligence. I'm scared. "

By contrast, Tech companies are mostly happy with the status quo, Because this status quo allows them to grab published works, To make their AI systems better mimic human behavior.

Sheila, America's top copyright official．Perlmutter (Shira Perlmutter) Hasn't taken a position yet. She told the Associated Press, She's listening to everyone, Because the copyright Office is weighing whether changes to copyright law are needed, To be adapted to produce compelling images, MUSICIANS, A new era of AI tools for video and text paragraphs.

Mr. Perlmutter said in an interview: "We have received it near1Ten thousand opinions. Every comment is read by a human, not a computer. I have read most of them myself. "

What's at stake here?

Perlmutter heads the U. S. Copyright Office, The bureau registered more than48Ten thousand Copyrights, Millions of individual works are involved. But there is a growing demand to register AI-generated works. So far, Applications for copyright registration for content entirely generated by machines have been rejected, Because the purpose of copyright law is to protect works created by humans.

however, Mr. Perlmutter asked, "When a human feeds content into an AI system and issues instructions to influence the output content, Is there a point where humans are involved enough to control the presentation elements of the output content, Thus it can be argued that humans contribute to authorship? "

This is a question that the Copyright Office is asking the public. Another bigger problem, It was also the creative industry that raised the issue with thousands of comments, What about copyrighted human works obtained from the Internet and other sources, These works are often used to train AI systems without permission or compensation.

in10Before the end of the initial comment solicitation period, The copyright Office got it9700Multiple opinions. The deadline for a new round of comments is12month6day. therewith, The Copyright Office will advise Congress and other agencies on whether changes are needed.

What are the artists saying?

"Kinship bond (Family Ties) " Actress and producer Justine．Bateman (Justine Bateman) To the Copyright Office, She's working on an AI model "Take in100Films of the year" And TV get upset, Such an approach could undermine the structure of the film industry, And displace most of the workforce.

Bateman claims: "In the opinion of many of us, It was the largest copyright infringement in American history. I sincerely hope that you will stop this theft. "

TV program "Poker face" The director of Lila．Zuckerman (Lilla Zuckerman) Also expressed concerns about artificial intelligence. She points out, She's in an industry that should be "Nothing but a plagiarism machine" The war on artificial intelligence.

Used to be Kelly．Underwood (Carrie Underwood) Wadgas．Brooks (Garth Brooks) Songwriting Nashville-based country songwriter Mark．Beeson (Marc Beeson) According to, The music industry is also under threat. Ai has the potential to do good, but "In some respects, It's like a gun. In the wrong hands, In the absence of any usage parameters, It could do irreparable damage to America's last true art form. "

Although most of the commenters are individuals, But their concerns are echoed by the big music publishers (Universal Music Group says the way AI is trained "Greed and lack of control" ) , Writers' groups as well as including "New York Times" And news organizations including the Associated Press responded.

Is this fair use?

Google, Microsoft andChatGPTmanufacturerOpenAIAnd other leading technology companies have told the Copyright Office, Their training of the AI model fits "Fair use" principle, The principle allows limited use of copyrighted material, For example for teaching, Research or transform copyrighted works into different content.

Facebook, InstagramandWhatsAppParent company ofMeta PlatformsSaid in a letter: "The artificial intelligence industry in the United States is partly built on this understanding, namely "Copyright law" The use of copyrighted material to train generative AI models is not prohibited. The goal of AI training is to recognize patterns through large amounts of content, Rather than extracting or copying individual works. "

So far, In explaining how copyright law should treat AI systems, The courts have mostly sided with tech companies. Last month, A federal judge in San Francisco has dismissed the first major lawsuit against an artificial intelligence image generator, But some cases were allowed to proceed, This is a failure of the visual artist.

2016years, The U. S. Supreme Court upheld the lower court's ruling, Dismissed authors' claims that Google's digitizing of millions of books and showing snippets of them to the public constituted copyright infringement. Most tech companies point to Google's successful defense as a precedent.

But a former law professor, To Courtney．Milan (Courtney Milan) Best-selling romance writer Heidi, writing under the pseudonym．Bond (Heidi Bond) consider, The comparison is flawed. Bond claims, She agrees "Fair use includes the right to learn from books" , But Google Books (Google Books) Obtained are legitimate copies held by libraries and institutions, And many AI developers are through "Outright piracy" To get the written work.

Perlmutter said, That's what the Copyright Office is trying to sort out.

Perlmutter points out: "The AI problem is different from the Google situation in some ways. Whether this difference is sufficient to preclude a fair use defence is the question at hand. " (Be compiled fromwww. usnews. com)

TRANSLATORS: Rason group　proofread: Liu Peng



　　disclaimer: This network reprint or compile the original articles are from the network, Does not represent the views of this website or confirm the authenticity of its content. If the source is mislabeled or the copyright of the article is involved, Please contact us, This website will be corrected in due course, delete, thank you.