Home > Insights > Blogs > Patent Billy Goat > So you want to protect your software: Tactics, options, and considerations post-Alice

So you want to protect your software: Tactics, options, and considerations post-Alice

Jason Schwent February 2, 2016

Since its first development, intellectual property law has struggled with how to best protect software. Initially, developers could only choose between copyright or trade secret protection to protect their innovative software. But, after the State Street Bank and Trust Co. v. Signature Financial Group, Inc. case in 1998, patent protection was available for certain software technology. But courts and the United States Patent and Trademark Office struggled with how to properly bring nebulous software innovations within patent protection. One of the chief concerns  whether software was even an eligible subject matter for patent protection.

This debate eventually culminated in the Alice Corp. v. CLS Bank Int’l case before the United States Supreme Court in 2014. In that case, the Supreme Court reasoned that the software patent at issue was invalid because it was directed to an impermissible “abstract idea.” Such abstract ideas, under prior Supreme Court precedent addressing patentable subject matter, were ineligible for patent protection. Using an analysis pulled from a number of the prior cases, the Supreme Court established a new test for determining whether a patent was directed to impermissible subject matter. This test, alternatively referred to as the Mayo (for an earlier Supreme Court case) or Alice test, operates as follows:

1. The Court is to first determine whether the claims of the patent at issue are directed to one of the defined protectable classes of patentable subject matter (i.e., a process, machine, manufacture, or composition of matter). If not, the analysis ends here.
2. If so, then the Court is to next determine whether the claims of the patent at issue are directed to a “law of nature,” “a natural phenomenon,” or “an abstract idea” — the so-called judicially recognized exceptions to the previously discussed classes of protectable subject matter. If not, then the claims are patentable and the analysis is over.
3. If so, then the final step is to determine whether the claims of the patent at issue recite additional elements that amount to “significantly more” than the law of nature, natural phenomenon, or abstract idea. If so, the claims are patentable. If not, then the claims are directed to impermissible subject matter.

The result of Alice, and the cases that have come since Alice, has been a drastic reduction in the patentability of software. Those areas of the Patent Office responding to patent applications directed to software areas of technology are reporting rejection rates, in some areas, of well over 80% (with some areas reporting 100% rejection rates). This raises an important question: How does one best protect software technology under existing intellectual property law?

What’s old is new again

One possibility is to return to the way software was treated prior to State Street. In those pre-State Street days, parties wishing to protect software were given a choice between either copyright protection or trade secret protection. But, because there are numerous ways to program software to achieve the same result and because neither protection would prevent “independent creation” by a third party, competitors are still allowed to simply start from scratch and write their own software to achieve the same result. So, while these options offered software developers some protection from outright theft, the protection afforded was minimal, at best.

Fight the good fight

While advocated by some of the judges in the decision, it is important to remember that the Supreme Court in Alice did not outright prohibit patent protection for software. It is still both theoretically and actually possible to obtain a software patent. While the number of software patents issued in 2015 dropped precipitously from those issued in 2014, software patents are still issuing. While this area of law is extremely volatile and evolving literally on a daily basis, some trends are beginning to emerge through the chaos. These trends indicate that though the Patent Office and Courts were initially very quick to strike down all software patents post-Alice, the pendulum has begun to swing ever so slightly in the opposite direction. Patents are issuing and Courts are upholding software patents.

So how does one obtain one of these rare software patents?

First, and perhaps foremost, it is critical to retain qualified counsel to assist you in this process who is up to date and knowledgeable on this topic. As noted above, this area of law is incredibly volatile right now. Proceeding without knowledgeable counsel is a recipe for sure rejection. Past guidelines in this area are not enough to find success. Too much has changed and too much is changing to rely on outdated practices.

Second, before attempting to acquire a patent on any software, it is necessary to do some pre-filing investigation. Not only should prior art be considered and addressed in the application, but one should anticipate receiving an Alice-based rejection and the argument you are going to make in response. Find cases where analogous technology has been upheld. Or, acquire expert opinions to bolster your positions. You cannot proceed under the hope that the Patent Office will not issue an Alice rejection. You must be prepared for it if you are to overcome it.

Third, in claiming your invention, where possible, emphasize physical results produced by your software. One of the key cases where software patentability was upheld was the case of Diamond v. Diehr. The claims of the patent in that case were directed to software which automatically calculated the curing time necessary for manufacturing rubber. But, beyond merely calculating the cure times for the particular rubber compounds, the software would physically shut off the ovens doing the curing at the prescribed time. It was this physical result which appeared to distinguish the Diehr patent from others conducting similar calculations. In terms of the Supreme Court’s Alice test, this additional step of physically stopping the curing process was the additional element that amounted to something “significantly more” than an abstract idea.

Fourth, when claiming your invention, it is important to emphasize the special nature of the computer at issue. Under the Supreme Court’s Alice test, a “generic computer” performing generic computer functions, like mathematical calculations, does not add something “significantly more” to the claimed abstract idea. However, if the computer system at issue is specialized in some way to perform only the patented function, then the Patent Office is more likely to allow the patent. This specialization does two things. It is evidence of additional elements amounting to something “significantly more” than an abstract idea. But it also serves as a hedge against the worry of preemption — the concern that the Patent Office (and especially Courts) have expressed that technology at issue, if patented, would preempt all innovation in the area. This policy concern is at the heart of the subject matter requirement — a concern that allowing patents on abstract ideas or natural phenomenon would impermissibly allow a single entity to “lock up” these fundamental scientific concepts. By emphasizing that the computer at issue is a special computer and not a generic computer, the concern that the technology at issue will preempt further innovation is lessened.

Finally, it is important to seek a patent only for what is actually invented. This may seem obvious, but common patent practice is to seek a patent having claims that are as broad as possible. One of the reasons software patents have been problematic is that many such patents use imprecise language to define their inventions — resulting in broad patent claims — but potentially covering technology not actually invented. Patent claims directed to a “network” for instance may have originally been directed to a pager network in the 1990s. But such claims could be applied against an internet network or cellular phone network today — networks which differ drastically from what was originally conceived. By focusing claims on what was actually invented (and not on other iterations that might possibly be covered), software patents may find more success with examiners trained to reject overly broad, overly vague patent claims.


Since the Supreme Court’s Alice decision, software patents have been an endangered species. But while their numbers have drastically declined, they are not yet extinct. Steps can be taken to still obtain patent protection for software innovations. Make sure you work with qualified patent counsel and follow the most recent guidelines and you may yet be able to patent your software technology.

Jason Schwent is a partner in Thompson Coburn’s Intellectual Property group. He is the editorial director of the Patent Billy Goat. You can find Jason on Google+Twitter, and reach him at (314) 552-6291 or jschwent@thompsoncoburn.com.