Micaela Ferreira’s Post

Now both AI music generator companies have responded to the lawsuits. Udio was far more measured than Suno but both basically laughed at the lawsuits and said, 'we're fair use, bro!' If there is a chance of any AI copyright lawsuit succeeding, it is related to diffusion models. Which both of these models are. I still do not think these models qualify as violating fair use either though. There is a mathematical equation, the Gaussian equation. The models try to solve for this equation. That allows for them to sample from the probability distribution, which creates their outputs. Where fair use violation, bro? I wrote a book specifically about all of this, breaking it down in every way and predicting how this would all go upfront. It is called, 'The Copyright Conundrum'. https://lnkd.in/g2i7g_ci

The record industry is suing AI companies over copyright. 🎶🤖 This harks back to the Napster era when they sued against streaming music, which ironically ended up saving the industry (thereby demonstrating a Luddite level of shortsightedness)😀📜 A key argument from AI companies is based on the premise that AI learns from data similarly to how humans do and then transforms the output. Looking forward to the court deciding on the question of “can machines learn similarly to humans?” ⚖️ For corporate leaders: ignore for now. This will take years to resolve and seems aimed at driving settlements and creating a market for AI-generated music. #AI #Copyright #MusicIndustry #ArtificialIntelligence #TechLaw #Innovation #FutureOfMusic #ai #lifesciences #ml

An important precedent may be set here (if it progresses to a hearing) with the main issue revolving around Suno and Udio using copyrighted music to train their AI models without obtaining proper licenses or compensating the artists/rightsholders. The infringement claim is seemingly pretty clear-cut but with AI not going anywhere, let's not go down the Napster path and seek to litigate against what may well be a future revenue stream. Hopefully it can be a more proactive approach in looking at how a licensing (and appropriate compensation) regime can be agreed upon between rightsholders and AI companies

The next stage of conflict between legacy content owners and upstart generative AI companies began yesterday. In this case the owners of rights to "human musical expression" -- major record labels representing the thousands of musicians and millions of songs -- sued two genAI companies for alleged copyright violations, asserting these companies used copyrighted material without consent to train their AI models, resulting in the unauthorized reproduction of songs. I find genAI-music to be obviously not human and a little off-putting, but it is here to stay; more to come on this subject. https://lnkd.in/dkX7wFnM

As widely covered in the media last week, our trade group the RIAA has announced legal action on behalf of members against generative AI companies Suno and Udio for infringement of US federal copyright law by each company on a massive scale. Suno and Udio allow users to generate digital music files that mimic genuine human created sound recordings in response to basic prompts, but as the cases make clear, these products can only work the way they do by copying vast quantities of sound recordings from artists across every genre, style, and era. At Sony Music, we are embracing AI and working with innovative start-ups that use AI ethically and responsibly. Like previous waves of new tech, innovation in AI should be based on business partnership between the creative and tech sectors that respects the value of artists and keep humans at the center of creativity.

Generative AI should not be a problem in the music industry as long as the models are trained correctly! With proper guidelines, Anthropic’s Claude chatbot recognized and upheld the copyright protection for “Paperback Writer”, refusing to recite all of the lyrics verbatim. “And speaking of Universal Music Group, they’ve certainly become known for being quite protective of their music rights, haven’t they?” - Claude, when prompted about being sued by UMG. Check out this MBW article including their full conversation with Claude for more info! #fsbmt #AI

💼 Anthropic & UMG Settle Landmark AI Copyright Case 💼 The legal clash between Anthropic and Universal Music Group has taken a decisive turn. A $95M settlement focuses on technical guardrails, allowing Anthropic’s AI models to evolve while preventing copyright-infringing outputs. What makes this settlement groundbreaking? 🎯 Guardrails Matter: The deal shifts attention away from the controversial use of copyrighted material during AI training to focus on preventing infringing outputs. ⚖️ Fair Use Debate Continues: Could this agreement influence future rulings on the legality of training AI models with copyrighted works? 🌍 Industry Impact: As AI systems integrate into industries like music, law, and tech, this case sets important precedents. Will this case redefine how intellectual property is managed in the AI era? 📖 Read more: https://lnkd.in/dg5rcKGe #AICompliance #IntellectualProperty #AnthropicVsUMG #CopyrightLaw #AIInnovation

AI and the Music Industry An AI-generated Drake song has sparked a debate about copyright and the future of music. The song "Heart on My Sleeve" features the voice of Drake rapping over a beat by Metro Boomin. It was uploaded to TikTok by an unknown user, @ghostwriter977, and quickly went viral. The song's origins are shrouded in mystery. Some speculate that it is Drake's publicity stunt, while others believe it is the work of a crypto-adjacent startup. Universal Music Group, Drake's record label, denied involvement but sent takedown notices to streaming services. As AI technology develops, creating realistic forgeries of popular artists will become increasingly easy. This raises several copyright questions. - Who owns the rights to an AI-generated song? - Should artists be compensated for using their voice and likeness in AI creations? The legal battle over "Heart on My Sleeve" is just beginning, but it could have far-reaching implications for the future of music. https://lnkd.in/g8kbfYCY

1/ Anthropic and major music publishers have reached an agreement prohibiting the AI assistant Claude from generating copyrighted song lyrics. 2/ The deal requires Anthropic to put safeguards in place and promptly address any reports of system failures from publishers. However, the underlying issue of whether Anthropic has the right to use copyrighted data, such as song lyrics, to train AI remains unresolved. 3/ OpenAI has announced plans for a "media manager" to allow rights holders to exclude their content from AI training, but has not provided any further updates on this matter since May 2024.

A surge in lawsuits against AI music generators like Suno and Udio underscores a massive moment in music... Major record labels, with coordination by the RIAA, are taking a firm stand against what they call "mass infringement" of copyrighted material. This legal battle emphasizes a fundamental question: Does using copyrighted music to train AI models constitute fair use? While these AI companies argue that it is innovation, the music industry says that unlicensed use threatens artists' brands and livelihoods. The outcome of these lawsuits will set a precedent for how AI interacts with intellectual property, not just in music but entertainment as a whole. I'm very intrigued to see how this plays out as someone who sees both the benefits and detriments of AI. What do you think? #ai #musicindustry #lawsuit #compliance https://lnkd.in/grvzd8Fu