The plan makes everything Obvious, which makes patents impossible.

Note: PHOSITA, pronounced “Fo-sit-ah,” means “Person Having Ordinary Skill In The Art”

[Edited for Grammar]

While there are significant downsides to excluding AI inventorship from patent law, these negative effects do not necessarily support recognizing AI as inventors under the current patent framework. If AI were to be recognized under the current patent framework, several major challenges would arise. First, patent eligibility is largely dependent on non-obviousness as measured against a “person having ordinary skill in the art” (PHOSITA). If the patent system simply adopted AI as an inventor under its current framework with no changes, this concept would be significantly undermined. As AI becomes more widespread, the concept of PHOSITA starts to break down. However, if the standard was modified to account for widespread use of AI, some have observed that everything would become obvious, because the inventing machines would be able to draw connections between all existing prior art, and, therefore, nothing would be patentable. Given AI’s capacity to analyze robust and expansive data sets, coupled with the availability of data through electronic sources like the internet, any possible connection or innovation would be obvious to every other PHOSITA aided by AI. Under the current patent framework, the growing use of AI also challenges the construction of the novelty requirement of patent law. The novelty test also relies on the hypothetical PHOSITA. Given even weak AI’s superior ability to analyze data, what would otherwise be inaccessible may in fact be accessible, changing what is novel in the context of patent eligibility. This flows from an analysis similar to the obviousness inquiry discussed above. Novelty depends on what references are accessible to the classic PHOSITA. As AI supplements human capabilities and overtakes humans in inventive roles, the scope of what is considered prior art expands significantly. It is important to note that both of these challenges exist independently of the recognition of AI inventorship. As the use of even Narrow AI expands, the concepts of novelty and non-obviousness are placed under increased tension. However, dealing with these general challenges to patent law with the continued development of technologies is beyond the scope of this Note. Nonetheless, these challenges are particularly acute in the context of AI inventorship. These tests are built into the statutory framework of US patent law. One major question that arises under the current framework as it relates to recognizing AI inventorship is whether a machine-created invention can ever satisfy the non-obviousness requirement. The requirement that an invention be non-obvious is patent law’s way of ensuring that the inventions that are granted patents are deserving of the protection that patents carry. Specifically, the non-obviousness requirement protects against granting patents to inventions that lack the element of creativity that patent law seeks to reward. AI-as-inventor ultimately puts this concept at risk because AI is not exercising creativity in the traditional sense. Much of the power that AI brings is drawing conclusions from large and unwieldy data sets. Ultimately, there is a real possibility that continued development of AI to ever-increasing levels will result in everything being obvious.