AccountabilityEthicsTransparencycs.CY
Montreal AI Ethics Institute's (MAIEI) Submission to the World Intellectual Property Organization (WIPO) Conversation on Intellectual Property (IP) and Artificial Intelligence (AI) Second Session
Allison Cohen, Abhishek Gupta
August 11, 2020
4 citations
Key Takeaways
Key Takeaways on AI and Intellectual Property
-
AI should not receive exclusive IP rights for inventions. The MAIEI argues that the case for granting AI systems their own intellectual property protections is weak at best, challenging the premise that AI "inventors" deserve legal ownership of their creations.
-
Human creators remain the true inventors. The paper emphasizes that humans—including developers, trainers, and those who directed the AI system—are the genuine inventors deserving IP protection, not the AI systems themselves, which are tools created by people.
-
IP protections for AI won't improve regulatory compliance. Contrary to arguments that IP protections could encourage responsible AI development, MAIEI finds that granting IP to AI systems is unlikely to actually ensure that AI developers follow ethical guidelines or regulatory requirements.
-
Granting AI IP rights creates significant harms. The submission identifies multiple negative consequences ("negative externalities") that would result from giving AI systems exclusive IP protections, though these harms require updating existing IP law to address properly.
-
IP law needs modernization for the AI era. Rather than extending traditional IP protections to AI systems, WIPO should develop new legal frameworks that account for AI-assisted and AI-generated works while keeping human creators at the center of IP rights.