No offense—innovation is high now, and resulting thickets turn all their internal links

The Al field already exhibits substantial innovation and disclosure under the current patent framework

The U.S. patent system exists to promote innovation and disclosure. Over the past decade, few fields have experienced a greater explosion in innovation than AI. There are new developments and advancements in AI technology all the time, as investment, research, and interest in the industry increased drastically over the past few years.18 AI development thrives everywhere, from universities and research institutions to startups and large companies. To remain globally competitive, the U.S. needs to keep investing in AI research and education.19

Notably, this innovation and growth in AI all happened under the current patent framework. AI development does not need extra incentive or stimulation. Thus, when considering any changes to how it evaluates AI patent applications, the Patent Office should carefully consider how those changes might impact this “Golden Age” of AI and whether they will ham­ per innovation. Easing the requirements for patentable AI inventions could have negative effects. In particular, the risk of permitting weak, overbroad patents, like those that plagued the patent system prior to Alice, could end up restricting downstream innovation in this field.

Additionally, the Patent Office should consider how any changes to the current patent framework might impact the high level of disclosure already in the AI development community. AI systems and tools are often distributed for free online under permissive licenses, whether they were developed by for-profit corporations, non-profit reach institutions, or a dedicated hobbyist. For example, TensorFlow,20 one of the most powerful and widely used libraries for machine learning in the world, is distributed un­ der the free and open source Apache 2.0 license,21 which includes an explicit patent grant. Patents naturally support the community’s open culture by promoting the disclosure of complex AI inventions that might otherwise remain secret in exchange for granting a temporary monopoly on the invention. However, patents without a proper enabling disclosure allow this monopoly without teaching the public how to make or use the invention. Granting these types of patents not only results in patents that fail under the statute, but also undercuts the open ideals of the community, whose work many in the field depend on and profit from.