This article may be too technical for most readers to understand.(October 2023) |
In US patent law, non-obviousness is one of the requirements that an invention must meet to qualify for patentability, codified as a part of Patent Act of 1952 as 35 U.S.C. §103. An invention is not obvious if a "person having ordinary skill in the art" (PHOSITA) would not know how to solve the problem at which the invention is directed by using exactly the same mechanism. Since the PHOSITA standard turned to be too ambiguous in practice, the U.S. Supreme Court provided later two more useful approaches which currently control the practical analysis of non-obviousness by patent examiners and courts: Graham et al. v. John Deere Co. of Kansas City et al., 383 U.S. 1 (1966) gives guidelines of what is "non-obvious", and KSR v. Teleflex (2006) gives guidelines of what is "obvious".
In the post-KSR (2006) patent legal practice in the US, the requirement for non-obviousness is often conflated with non-predictability. It is easier to find non-obviousness in non-predictable arts (such as pharmacology), and more difficult in predictable arts (such as business methods). Another contentious area is contradictions between non-obviousness and patentable subject matter requirements: whereas non-obviousness can be easily met, when inventor discovers something new and unexpected (like the discovery of the DNA segments corresponding to the fetus's father in the blood of pregnant women in Ariosa v. Sequenom),[1] a patent will be denied, as not meeting statutory subject matter requirement, "if the APPLICATION [of this discovery] merely relies upon elements already known in the art."[2] (see also Mayo Collaborative Services v. Prometheus Laboratories, Inc. and Association for Molecular Pathology v. Myriad Genetics, Inc.) The alleged rationale for this dichotomy is to prevent broad patent claims that would inhibit rather than "promote the Progress of Science", as required by the Copyright Clause of the US Constitution.[3]