In earlier days it was often claimed that the conditions for applying for a software patent ("computer-implemented inventions") were less complicated in the US than in Europe. On 19 June 2014, however, the American Supreme Court passed a crucially important judgement in the case of Alice vs CLS Bank. In short, since that moment a two-step assessment must be performed before a software invention can be patented in the US.
First of all, it must be determined whether the claims in question are aimed at a concept that has been excluded from patenting. Examples of such concepts are abstract ideas (which sometimes include certain abstraction levels of software), natural laws and natural phenomena. If the claims are focused on concepts in these categories, it must be determined whether the elements of the claim, regarded individually and as an orderly combination, transform the nature of the claim into an application for which a patent can be submitted.
The second determining factor is to establish if the elements of the claim can be seen as an inventive concept, which is sufficiently concrete to transform the claimed abstract idea into an application for which a patent can be applied; those elements must help in obtaining something that is "significantly more" than a non-admissible abstract idea. In other words, additional features need to be included with the claim in order to ensure the claim goes beyond being merely an abstract idea and not to simply claim to execute the abstract idea using an everyday (generic) computer.
One example of an American patent that didn't make it thanks to the new two-step assessment follows from a ruling by the Federal Circuit in the case of Internet Patents vs Active Network on 23 June 2015. This case concerned using a conventional navigation functionality (Forward and Back) of a web browser without data loss in an online application consisting of dynamically generated web pages. According to the Federal Circuit, this was material which could not be patented due to the fact that this concerned a generic data collection process in a specific technological environment. To put it differently, the claims were considered to be aimed at the abstract idea of maintaining the "computer state", without indicating specific activities to reach that objective.
A second example of an American patent that didn't see the light of day is the case of Mortgage Grader vs First Choice Loan Services of 20 January 2016. In this case, the Federal Circuit ruled that the claims were aimed at systems and methods to help people in obtaining loans — the abstract idea of "anonymous loan shopping" —, because the claims did not mention anything more than the collection of information to set up a "credit grading" in order to make anonymous loan shopping possible, and simply added generic computer components, such as an interface, a network and a database.
Thanks to the two-step assessment of the Alice case, it looked like it would become significantly more difficult to apply for a patent for software based on a more application-oriented level and under abstraction of the specific implementation features.
An important twist came about, however, in the case of Enfish vs Microsoft on 12 May 2016. This case concerned two patents aimed at an innovative logical model for a computer database, in which all data entities are housed in a single table and wherein column definitions were provided by rows inside that same table, as a "self-referring" feature of the database. Although the initial ruling (before the District Court) determined that this concerned the abstract idea of saving, organising and retrieving memory values inside a logical table, and that the content could therefore not be protected, the Federal Circuit was of the opinion in the second instance that this was indeed an invention for which a patent application could be submitted.
The Federal Circuit judged that the actual focus of the claim was on the improvement of computer functionality itself and not on other economic or other tasks for which a computer is used in its ordinary capacity. In this case, the self-referring table was in fact a specific type of data structure designed to improve the manner in which a computer stores and retrieves data in its memory. Furthermore, the Federal Circuit was of the opinion that the patents in question contained specific technical wordings with regard to the features of the self-referring table. The Federal Circuit also ruled that the mere possibility that the invention could be carried out on a regular (general-purpose) computer instead of a specially modified machine did not necessarily make the claims invalid.
To summarise, the Enfish analysis poses the question whether the focus of the claims is on a specific improvement in the functioning of a computer, or rather on a process which is actually an abstract idea that is executed using a computer.
Based on later judgements by the Federal Circuit, the same is more or less apparent: a claim aimed at carrying out an abstract idea with the help of conventional or generic technology in a familiar environment is not valid (in re TLI Communications dated 17 May 2016). Conversely, a technology-based solution (so not an abstract idea-based solution with generic technology applied in a conventional manner) to resolve a specific problem inside a specific application is eligible (BASCOM Global Internet Services vs AT&T Mobility of 27 June 2016).
These rulings mean that the American patent practice for software has moved a step closer to the European system. For years the European patent practice has based its rulings on the premise that a claim containing no technical elements (for example a purely economic idea or a purely mathematical process, in which no technological or physical features are present) is inadmissible. A claim must in any case contain at least a technical element and must also be inventive: a non-obvious solution to a technical problem.