US Software Patents
Yes, Virginia, software can be patented.
The 1998-07-23 State Street Bank v. Signature Financial Group CAFC decision opened the door (some would say “floodgates”) to business method patents. The current backlog at the USPTO is directly tied to the State Street decision.
The 2012-06-28 Bilski v. Kappos Supreme Court decision said that the correct (to-be-determined) standard for patent subject matter eligibility is somewhere between the standard set in the CAFC’s 1998 State Street decision (“useful, concrete and tangible result” test is too broad) and the CAFC’s 2008 Bilski decision (“machine-or-transformation” test is too narrow). In other words, the Supreme Court essentially reset software patent law to 1982.
KSR and Bilski give the USPTO – which is already predisposed to rejecting patent applications – more ammunition to make rejections. Regardless of how you interpret these cases, the USPTO has a track record of publishing “guidelines” for how examiners should interpret and apply the law. These guidelines are rarely favorable to applicants.
- On 2007-10-10 (and 2010-09-01), the USPTO published guidelines for KSR.
- On 2010-06-28, the USPTO published guidelines for Bilski.
In light of KSR, if your invention consists of combinations of elements that exist in the prior art (either literally or equivalently), then it will be extremely difficult – if not impossible – to secure a patent. Conversely, if your invention has the “nugget of newness” (Clocktower Law’s term) not found in the prior art, then it has a better chance of being patented.
In light of Bilski, if your invention does not strictly require a computer to be carried out, then you probably won’t be able to patent it. Conversely, the more complex and product-ized your software invention, the more likely you’ll be able to patent it.
The bottom line is that software has been patentable for over a generation and remains patentable today, in large part because courts are reluctant to change established laws.