G1/19: Are simulations inventive?
Posted in | Aug 28, 2019
This series of articles considers referrals pending before the Enlarged Board of Appeal of the European Patent Office (EPO). Whilst the outcome of these referrals may be some months or years away, the opinions issued by the Enlarged Board of Appeal may fundamentally change the way the law is interpreted. In this article, we look at referral G1/19.
G1/19, a referral made to the Enlarged Board in February 2019, relates to whether computer simulations can be classed as inventions under the European Patent Convention (EPC).
The applicant (now appellant) applied for a patent relating to computer simulations of crowd movement. The simulations were said to be more advanced than the state of the art at the time, as they included concepts of personal space, frustration and inconvenience for the computer simulated people in the crowd.
However, the application was refused (in several amended forms) for not being inventive.
The appellant appealed this decision, and the appeal has resulted in this referral to the Enlarged Board which is seeking to clarify whether and under what circumstances a computer-implemented simulation can give rise to a technical effect for assessment of inventive step.
As usual, the referral takes the form of a series of questions to the Enlarged Board. The questions do not reflect the actual case from which the referral is made but are framed more generally at the point of law in question.
The Board of Appeal has referred three questions to the Enlarged Board for a decision. These are:
1. In the assessment of inventive step, can the computer-implemented simulation of a technical system or process solve a technical problem by producing a technical effect which goes beyond the simulation's implementation on a computer, if the computer-implemented simulation is claimed as such?
Traditionally, the EPO sees processes using mathematical calculations performed by computers as non-technical features of the invention, so they cannot be used to support inventive step. This often results in rejection of claims relying on these features as not patentable under Article 52(2)(c) EPC.
However, if a mathematical method is used to monitor, change or oversee a real-world physical process, this overcomes the non-technical feature rejection, as it has a technical effect on the process.
With that in mind, the question effectively becomes:
Should the EPO allow computer-implemented simulations which solve a technical problem in a way that provides technical benefits outside of the simulation, to be classed as technical features?
This is a particularly important principle, as the recent surge in AI related inventions means that this is an issue that will arise more and more often in the near future.
2. If the answer to the first question is yes, what are the relevant criteria for assessing whether a computer-implemented simulation claimed as such solves a technical problem? In particular, is it a sufficient condition that the simulation is based, at least in part, on technical principles underlying the simulated system or process?
The first question seeks to determine whether as a matter of principle computer simulations when claimed as such can give rise to a technical effect. If the Enlarged Board says that they can, these follow up questions are intended to define the circumstances under which a technical effect can arise. In other words, what boundaries must be drawn to ensure that this principle is not abused by applicants? Must the simulation be of a sufficiently technical process? The answer to this (hypothetical) question is not simple, and may well be the subject of further case law if the Enlarged Board decides computer-implemented simulations can be used to support inventive step.
3. What are the answers to the first and second questions if the computer-implemented simulation is claimed as part of a design process, in particular for verifying a design?
The final question referred to the Enlarged Board seeks to establish the position when a computer-implemented simulation is claimed as part of some larger process, and in particular a design process. To what extent can using a computer-implemented simulation contribute to the inventive step of a larger design process? Is simply using a computer-implemented simulation enough to provide inventiveness to such a process?
If so, his could become very important if AI and machine learning become prevalent in industry (as expected by many people) as it may provide a loop-hole for gaining patent protection for inventions relating to AI and machine learning.
This case has the potential to have enormous repercussions for practice at the EPO. A decision that computer-implemented simulations can be classed as inventive would mean many claims that would previously be seen as unpatentable or unlikely to be allowed could suddenly become patentable inventions in their own right.
With AI developing at an extremely fast rate, a decision that computer-implemented simulations can be used to support inventive step could open the flood gates to many new patent applications, and raise entirely new points of law in this area. The Board of Appeals have tried to mitigate this with their second question, but as with all ground breaking decisions on points of law, only time (and the case law) will provide answers on the nuances of computer simulation inventions.
If the Enlarged Board decide computer-implemented simulations cannot be used to support inventiveness, very little will change in terms of practice at present. However, this would be the first Enlarged Board decision (and possibly not the last) that would provide hurdles to patenting AI and machine learning inventions. In these circumstances, there would undoubtedly be further appeals and referrals in order to establish the extent to which AI inventions are allowable under the EPC.
If you have any questions about this decision or would like general advice on practice before the EPO, please get in touch to speak to one of our attorneys.