This week, Digital Future Daily will dedicate five days to the fast-moving landscape of generative AI and the growing conversation about how and whether to regulate it — from pop culture, to China to the U.S. Congress. And the Grammy goes to… an AI music generator? The breakout hit of the spring is “Heart on My Sleeve,” a track that sounds just like a collaboration between musicians Drake and The Weeknd — two mega-popular artists who didn't record any of the song themselves. The track relies purely on AI-generated imitations of their voices, posted online by a pseudonymous TikTok user, and since it went viral last weekend it’s been heard millions of times. It has also generated takedown notices and a statement from The Weeknd’s label, Universal Music Group, slamming AI-powered copyright infringement and calling for users of the technology to get on the right “side of history.” That makes it another flashpoint in a larger conflict now brewing over the future of intellectual property in the age of AI — and also a kind of early warning about the growth of artificially generated culture. From courtrooms to Coachella, that fight has kicked off in earnest in recent weeks, with pressure campaigns and legal battles taking shape across the U.S. It's pitting artists and media conglomerates against AI companies and their users, with tech platforms often caught in the middle. The outcome will have profound consequences for the future of IP, a pillar of post-industrial economies. AI is already roiling the music industry. “Heart on My Sleeve” may be among the most successful examples of machine-generated media, but it’s just a part of a flood of AI-derived images and sounds that has washed across the Internet in recent months. That flood hints at a version of our digital future in which each of us lives in a fully personalized entertainment bubble, with videos and songs pumped out by AI in seconds, just for us. That's still a ways off, but there's a much more immediate change already afoot in which AI composes songs, or renders art, or writes stories which might not be quite human-level, but are good and cheap enough to start displacing human work. That version of the future has already begun, and the backlash is already underway. Last month at SXSW in Austin, a coalition heavy on unions and media trade groups launched the Human Artistry Campaign to push back against encroachment by machine-generated media. Mitch Glazier, chairman of the Recording Industry Association of America, a member group, told DFD that creative industries are speeding up their political organizing efforts in response to AI’s rapid evolution. In addition to musicians and record labels, groups representing photographers, literary agents, actors, writers, and athletes have already joined the campaign. “It’s an amazing group of creative industries that has signed onto these principles in a very short amount of time,” he said. For firms like Google — actually, for Google in particular — as Nilay Patel noted in the Verge last week, this fight presents an especially acute dilemma. The company has to mollify copyright holders that use services like YouTube, which took down the fake Drake-Weeknd song. But it also has invested heavily in the underlying AI technology, meaning it has an interest in establishing as much legal leeway as possible for the uses of generative AI. Some artists are breaking ranks to embrace AI-powered appropriation of their work. Last night, the musician Grimes, known in the tech world for her brief marriage to Elon Musk, tweeted a standing offer to split the royalties 50-50 with anyone who makes AI-generated music that uses her voice. AI-generated music seeped into talk among industry types at both Miami Music Week last month and at the just-concluded Coachella Music Festival in Southern California, according to Nicholas Saady, a tech-focused attorney at Pryor Cashman, a firm with an active music litigation practice. “It’s one of the hottest topics in music right now,” he tells DFD. Images are another front in this battle. As with music, legal fights over AI images and music fall into two broad categories, inputs and outputs. While an AI output that sounds just like Drake raises one set of issues, another broad set of issues is raised by questions about whether large language models — the platforms behind tools like ChatGPT — can legally use copyrighted works as training material for their algorithms in the first place. (The way they work right now, their training depends on inhaling a huge corpus of human writing and art, much of which is copyrighted.) One early test case is making its way through federal court in Northern California. On Tuesday, Stability AI and Midjourney, companies that offer AI image generators, filed responses to a proposed class action suit by visual artists who allege the models infringed their copyrights. The companies are seeking the dismissal of the lawsuit, arguing the plaintiffs failed to identify any copyrighted works used as inputs, or any outputs that resemble copyrighted works. The AI pushback is only getting started. Glazier said most industry participants believe that current law already restricts AI companies from training their models on copyrighted works and from producing imitations of protected works and artists. At least, that is, “if applied correctly by the courts.” He said creative industries were closely watching a suit filed in February in federal court in Delaware by Getty Images alleging that Stability AI unlawfully used its photos to train its LLM. Members of the Human Artistry Campaign are likely to file briefs in support of Getty’s case, he said. There’s also a Plan B. If the courts don’t rule on behalf of copyright holders and artists, Glazier said, “legislation might be necessary.” But don’t expect the creative fields to call for an outright ban. In pushing back on generative AI, everyone from the RIAA to the Human Artistry Campaign to Getty has also talked about the potential for artificial intelligence to enhance the artistic process: as long as it’s done within their preferred legal framework. So, however these fights play out in courtrooms and legislatures, society will likely end up looking to artists, just as much as policymakers, in the search for an appealing balance between humans and machines.
| 
