Dibachi suggests that patenting AI-related inventions is challenging in the U.S., as it requires proving that the innovation was a product of human invention and not AI. Copyrighting AI IP is also problematic, as the Copyright Office only allows protection for the human-generated parts of AI IP systems. She concludes that trade secret law, which hinges on the confidentiality of the information, emerges as a more fitting form of protection for AI innovations.
Key takeaways:
- AI-generated intellectual property (IP) is treated differently, both legally and practically, than human-generated content, and understanding these differences is crucial for protecting AI IP.
- Patenting AI-related inventions is challenging in the U.S as it requires proving that the AI-related innovation was the product of human invention and not the product of AI itself.
- Copyrighting AI IP is also complex, with copyright protection only extended to the human-generated parts of AI IP systems and not the AI-generated portions.
- Trade secret law emerges as a more fitting form of protection for AI innovations, as it hinges on the confidentiality of the information rather than the nature of its creation.