Max Planck Institute for Innovation and Competition, Munich, Room 313
Considering that both patents and trade secrets can promote knowledge diffusion via licensing, we compare patents and secrets on the way they facilitate diffusion through non-exclusive licensing. This comparison is important because AI algorithms are not patentable. Patents differ from trade secrets on the easiness of imitation when licensed to many downstream producers. In view of this, we argue that secrets are better in defending one’s monopoly under the threat of imitation, inducing market failure on the diffusion of AI. We also find a non-linear relationship between patent strength and the diffusion of ideas, i.e., patents of intermediate patent strength are best as diffusion instruments.
Contact: Dr. Fabian Gaessler