Founder @TREAT AI - Open-Source AI for Detecting Triggering Content in Movies & Shows | Perplexity AI Business Fellow | SWE Virtual Exp Intern @JPMorganChase | BITS Pilani CS 27' | GGSIPU 28'
Let me ask you a fun question 👀 If you created content using AI, who holds the copyright to it? Is it you? Is the AI platform you made it on? Is it nobody? Or is it the creator of the data the AI was trained on? Well, whatever answer you do have, chances are, you are probably wrong, and I'm worried this is about to become a 𝐛𝐢𝐠 𝐩𝐫𝐨𝐛𝐥𝐞𝐦. The question of who owns AI-generated content is more than just an academic curiosity—it’s a growing concern that could impact 𝐜𝐫𝐞𝐚𝐭𝐨𝐫𝐬, 𝐛𝐮𝐬𝐢𝐧𝐞𝐬𝐬𝐞𝐬, 𝐩𝐥𝐚𝐭𝐟𝐨𝐫𝐦𝐬, 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐜𝐫𝐞𝐚𝐭𝐢𝐯𝐞 𝐢𝐧𝐝𝐮𝐬𝐭𝐫𝐲. In my latest article, I explore the murky waters of AI and copyright law, examining who might truly hold ownership and the potential consequences for misuse and claims. The lack of clarity has real implications, from the misuse of Content ID systems to disputes over derivative works that can affect musicians, artists, and content creators alike. 𝐈𝐬 𝐢𝐭 𝐭𝐢𝐦𝐞 𝐟𝐨𝐫 𝐚𝐧 𝐨𝐯𝐞𝐫𝐡𝐚𝐮𝐥 𝐨𝐟 𝐜𝐨𝐩𝐲𝐫𝐢𝐠𝐡𝐭 𝐥𝐚𝐰𝐬 to keep up with the rapidly evolving world of AI? Share your thoughts below! 💭 #AI #DataEthics #Future #technology #Law
Information Research Professional
2moFrom an IP Standpoint (and, BTW, I'm not a lawyer), it's known that while new, novel, and non-obvious work is patentable it doesn't extend to software (perhaps in the USA, but I believe that's fallen out of favor). So, while copyright could be claimed on AI Software its' patentability is probably suspect. (I'll leave that to the lawyers ;-) New intellectual property doesn't exist in a vacuum, it's a collection of small, individual advancements built on a history of small, individual advancements (and, in some cases, it can break new ground, say, by a Tesla or Edison). When someone claims a copyright or receives a patent they've (presumably) successfully shown distinction from the known art, but can be challenged anytime and overturned by Prior Art. I suspect even if two people were to ask the same task of AI, even to the same AI Engine pre-trained the same way, each would generate a new result having some distinction from the other. And, perhaps, each could claim copyright and be subject to the same prior art challenge. The problem I see is that unlike traditional IP the potential database of AI-generated work product could be infinite, making any challenge (ahem) "challenging". Perhaps AI would help ;-) Just my 2 cents ;-)