Canada Software Patents – What Does MOPOP Say?

The Canadian Intellectual Property Office (CIPO) issues a Manual of Patent Office Practice (MOPOP) to guide applicants on the application procedure. The latest edition of MOPOP has a chapter dedicated to clarify the patentability of Computer Implemented Inventions. 

The first point emphasised by MOPOP is that Schlumberger, though the only authority on software patents, did not decide that computer implemented inventions are not patentable, but instead held that the introduction of a computer aspect to to the invention does not, by itself, make an invention any more or less patentable, and that a computer implemented invention would have to satisfy all the regular patentability requirements for the application to succeed and must fall within one of the invention categories mentioned in Section 2 of the Patent Act.

On issues of subject matter MOPOP makes it clear that a computer implemented process will only qualify as patentable subject matter upon the satisfaction of two conditions. The first is that the process must be an act or a series of acts performed by some physical agent, upon some physical object and produce in such an object some change either in character or condition.  The second condition is that the process must produce any essentially economic result relating to trade, industry or commerce. 

On the application of Subsection 27(8), MOPOP says that a claim to a method consisting only calculations made in accordance to certain formulae is excluded from patentability, because of its lack of any act performed by a physical agent upon some physical object and producing such a change of character or condition in such object. The same method will also not qualify because it does not produce an essentially economic result relating to trade, industry or commerce.

MOPOP states that there are three possible qualifying type of claims for which a computer implemented invention may be successfully patented:

  1. Art or Process (Method) Claims – These are claims for a process that defines a series of operations taking place when a computer program is run. A claim made in this form must satisfy the process requirements mentioned earlier.
  2. Machine (Apparatus) Claims – These are claims for a computer which has been configured with a novel compute program. The claims should be drafted as “means for” expressions and must involve the same level of functional step detail expected of the process equivalent claim for the same invention.
  3. Manufacture (Products or Computer Media Embodying Code) Claims – These are claims for computer readable memory devices that contain instructions to be executed by a processing system. The process to be executed by the ‘reader’ device must satisfy the requirement of a qualifying patent as specified earlier.

MOPOP also provides examples of how a qualifying claim would be drafted for each of these as well as examples of what would not be accepted as qualifying subject matter.

The MOPOP guidelines themselves seem to be very consistent with the case law and were updated before the more recent liberal decisions taken by BAP.

Should Software Patents Be Granted at All?

Software Patents
(Photo Credits: Leonard Low)

Software Patents introduce unique problems to the patent system. The main goals of any patent system is to create an incentive and a reward for inventors who create helpful inventions. There is no doubt that the computer software is one of the most important industries in today’s economy. However, they argument as to whether or not they should be patented at all is still hotly debated. 

Arguments Against Software Patents

There are several grounds for refusing software patents. The first one is that computer software is already protected by copyright. The protection of copyright is automatically granted without any formalities and it grants to the author of the software with the right to prohibit others from copying the way the software is expressed – i.e. the literal copying of the program code. While it is true that copyright’s protection differs from that of patent protection, in that a person can replicate the concept of the software without infringing the copyright if he does not copy the same code, any greater protection would be essence go beyond the expression and into the algorithm and idea behind the software and that should never be subject to patent or copyright protection as the free flow of ideas is necessary for any society to thrive.

Software patents do not easily fit into the patent system because they do not give society much value in return for the patent bargain. The patent system runs on the basis that an inventor would be granted protection for his novel invention upon the disclosure of the workings on the invention. The inventor gets monopoly over the invention and society gets to learn how the invention works through the disclosure. Satisfying this bargain is hard to achieve in the case of software patents because the applicant is never required to disclose the source code of the software, and even if they do, it is impossible for a patent examiner to determine whether this code is sufficient to rework the invention due the massive amount of the code.  This makes the software patent database available to the public pointless as they cannot learn anything about the invention by analysing the information provided and therefore the bargain is not really balanced.

Another problem with software patents is that they the length of a patent goes way beyond any expected market life of any software. The patent system was invented for “slow” industries for which the 20 year patent term allowed a company to generate a return on investment that covers its research and development costs. It is also likely for the patent subject to be still of commercial value even after the expiry of the term so that competitors can use it and built upon it. This could be not said about software patents as no software would have any value after a 20 years period and it would be very unlikely for society to derive any benefit from the the patent subject after the expiry of the term.

The earlier point is amplified by the fact that a patent application could taken up to 4 years for the patent office to examine it and grant a patent. This does not exceed the average market life of a software, but also subjects other software developers to the danger of unintentional patent infringement as they would not know about the existence even after the software becomes obsolete.

Many also believe that software patents are not necessarily at all to create an incentive for innovating in the software industry. This is apparent from the success of the “open source movement” and the success of a number of business institutions in creating profitable products and services built using open source software. In contrast, patented software is accused of being a cause for the lack of inoperability between different software and hardware and the emergernce of incompatible computing standards.

(Software patents are also very difficult to examine due to the unavailability of a sufficient database, complexity of software code, and the fact that source code is not submitted).

Arguments in Favour of Software Patents

As the industry changes and the world moves away from the ‘mechanical’ industries into digital industries it becomse essential for the patent system to adapt to these changes and provide incentive and reward for influential industries which are of the greatest important to society. 

TRIPS states that patent protection should be available for inventions in all fields of technology and the software industry is clearly an industry worth protecting. 

Though copyright may protect the expression of a computer code, this protection does not extend the way in which this code was applied in industry, and that is the protection granted by the patent system. Copyright is effective in protecting against consumer piracy, but it does not protect against the duplication of the invention by a competitor. As the investment in R&D in the software industry grows exponentially, it becomes vital for the state to provide those who invest in this industry of a method for protecting their investment. Trade secrecy is a viable solution only for a small number of business, but it is hard to maintain in large scale operations and in circumstances where the inner workings of the invention can be protected through reverse-engineering.

The available of patents is also helpful for small and medium enterprises who cannot compete with larger companies if their inventions could be easily duplicated without the protection of patents. The successful registration of SMEs is also used by these institutions as a marketing tool to attract capital which is necessary for their businesses to grow and expand.

I think that both sides of the debate have valid points. The software industry should have some protection, but the patent system clearly does not easily accommodate it. At times I think that the solution would be in the creation of a new sui generis type of protection for computer software that involves quicker examination, shorter protection term, and has protections and defences for third party in a manner that serves the software industry as it stands today.

Software Patents in Canada – Changing Position of CIPO (3)

The Canadian Intellectual Property Office (CIPO) is the body responsible for granting patents in Canada, it also issues the Manual of Patent Office Practice (MOPOP) which is used as a guideline for examiners and applicants. Though the courts opinion regarding software patents does not signal in its favour, CIPO has made a number of recent decisions allowing software patents even though it used to also have a strict position against them in the past. The latest edition of MOPOP also has a chapter explaining the requirements for a software patent to succeed.

CIPO Software Patent Decisions

After the Schlumberger court decision, CIPO took a strict position against software patents until 1998 when the Patent Appeal Board decided in favour of two software patent applications. The first is Re Motorala Inc. Patent Application No 2, 085, 228 which involved a device and a method for evaluating exponentials. The claims in the application were originally worded in the format of method claims, later changed in the prosecution stage into apparatus claims. The ‘invention’ was made up of a special software that was coupled with a Read-Only Memory (ROM). CIPO held this to be a qualify invention because of the inclusion of the ROM as a specific part required to run the software:

[T]he Applicant has invented a device which is specifically adapted to carry out the method of solving the algorithm which the Applicant has developed. This device, while it does contain many means-plus-function statements, also includes at least one specific piece of computer hardware which is a real physical element. As a result, the Board believes that the claims of this application go beyond being directed to a mere scientific principle or abstract theorem. The Applicant is not seeking to exclude others from using the algorithm itself but is seeking to exclude others from using the specific device which is claimed.

This decision is a departure from Schlumberger as the question of where the inventiveness lies was never asked, if it were it would be clear that the inventiveness here was in the software and not the hardware nor the combination of both. While the existence of a hardware is necessary for the software to run, this piece of software is generic as any computer program will have to be stored in the ROM of a computer for it to operate, and in fact, the claims of the patent would be infringed if only the software was copied and then was used on any computer as the invention would consequently be worked that way (A5).

While this new approach might be favourable by applicants, it is a departure from case law and might as well be rejected by the court if a case was ever to reach the courts. CIPO held an identical decision also in favour of the applicant in the decision regarding the Re Motorola Inc. Patent Application No. 2,047,731 where the invention related to an apparatus for carrying out a convergence algorithm for use with selected convergence rates which gives improved efficiency on math computation of J-th roots. The software was coupled with a generic piece of hardware and consequently accepted by PAB as a qualifying subject matter that is not a “mere scientific principle or abstract theorem”. It is clear from these two decisions that ANY software could satisfy subject matter requirements by simply coupling it with a generic piece of hardware.

While the Motorala decisions show the general change at CIPO in favour of software patents, other decisions seem to also soften the impact of other cases, such as Progressive Games, which could have an impact on software patents through the exception of games, business methods, and professional skills.

The first example of these new decision is Application No. 2,298,467 which involved a system for facilitating the exchange of data between two computers for the purpose of trading diamonds. The application was rejected by the examiner of several grounds including the fact that this involved a professional skill and a business method, both of which are unpatentable subject matter. The PAB rejected both of these arguments saying that “Professional skill that falls outside the scope of patent protection involves a step in a claimed method which is carried out by a human and which relies on the intelligence and reasoning of the human to make a judgement.”

The PAB avoided discussing the issue on whether or not business methods were patentable at all but stating the the claims fell within the definition of a “machine” and not a business method. This way the other rejections made by the examiner on the grounds of business methods were rebutted.

This of course is a much narrower definition of professional skill than that described in Lawson and requires an actual participation of a human person in the process for it to be disqualified.

Another decision with a similar outcome is the one relating to patent Application No. 2,119,921 which involved an improvement to a computerised stock trading system for which one of the claims was for displaying and communicating orders to an order entry system of a stock exchange computer. The examiner rejected the application on a number of grounds and one of these was on subject matter on the basis that it was that it related to the automation of features taught by prior art. The PAB rebutted the rejections of the examiner and said that the examiner confused subject matter issues with obviousness when obviousness should not form the basis of a subject matter consideration.

The PAB has also recently showed its willingness to reject the court’s analysis in Progressive Games in the decision of Patent Application 2,220,378 in which the PAB had to decide on the patentability of an application involving systems and methods for group billing in cellular telephone plans. The examiner rejected the application on the grounds of subject matter on the basis that it did “not substantially modify the art of the billing system, nor [does it] create a new or improved billing system” and that it does not “amount to a contribution or addition to the cumulative wisdom of billing systems” – which is the same ground for refusing a process according to Progressive Games. 

The PAB rebutted the rejection of the examiner stating that subject matter should be examined on its own without taking into consideration issues of novelty, obviousness, and utility and that examination should only decide on whether or not the invention can fall within one of the patentable categories of section 2, namely an art, a process, a machine, a manufacture, or a composition of matter.

This is a very logical explanation of what should be done by CIPO, but CIPO does not have the legal authority to overrule a court decision, which in this instance was clearly disregarded.  It is clear that CIPO is now more willing that ever to allow a software patent to successfully pass on the grounds of subject matter, but that does not meant the court will have the same opinion as CIPO in matters in which the court previously held otherwise.

Software Patents in Canada – The Position of the Courts (2)

In addition to the possibility of having an software patent application rejected on the grounds that it was a mere mathematical formula or abstract theorem in accordance with Section 27(8) and Schlumberger, the application might also be rejected for failing to satisfy the subject matter requirements for not being ‘art’ in accordance with section 2.

A computer program could be in essence one relating to a business method or a game, both of which have been rejected by the courts in Canada as qualifying subject matter. A business method or a game could theoretically fall under the definition of the term ‘art’ mentioned as a type of invention in section 2 of the act. In the case of Shell Oil Co v Commissioner of Patents 67 CPR (2d) 1 the court held that the “word “art” in the context of the definition must be given its general connotation of “learning or knowledge” as commonly used in expressions such as “the state of the art” or the “prior art”“.

This definition had been restricted by the courts in previous decisions to state that the term ‘art’ that do not include “professional skills”. In Lawson v Commissioner of Patents (190). 62 CPR 101 (Ex. Ct.) the court held that a method of describing and laying out parcels of land in a unique shape was not patentable as failing within the skill of a solicitor and conveyancer and that of a planning consultant and surveyor. It therefore rejected the application on the grounds that it involved a professional field and that did not fall within the meaning of section 2.

The case of Tennessee Eastman Co. v Commissioner of Patents (1970) 62  CPR 117 (Ex. Ct.) also had a decision similar to Lawson in that the court held that methods of medical or surgical treatment were also excluded because they were essentially professional skills.

It is worth noting that Lawson also defined “art”:

An art or operation is an act or series of acts performed by some physical agent upon some physical object and producing in such object some change either of character or of condition. It is abstract in that, it is capable of contemplation of the mind. It is concrete in that it consists in the application of physical agents to physical objects and is then apparent to the senses in connection with some tangible object or instrument.

The definition makes it clear that the definition of “art” requires the existence of “physical” tangible aspects for it to qualify which may sound like a very restrictive definition, however, the court later said in the case of Shell Oil Co v Commissioner of Patents (1983) 67 CPR (2d) 1 (SCC) that “‘art’ was a word of very wide connotation and was not to be confined to new processes or products or manufacturing techniques but extended as well to new and innovative methods of applying skill or knowledge provided they produced effects or results commercially useful to the public”. 

The Shell definition is a much wider one than that of Lawson, but because of the way it was worded, it was understood by the court in the case of Progressive Games, Inc v Commissioner of Patents (1999) 3 CPR (4th) 517 (FCTD) to mean that for a process to be considered an ‘art’ it must:

  1. not be a disembodied idea but has a method of practical application;
  2. be a new and innovative method of applying skill or knowledge; and
  3. have a result or effect that is commercially useful.

The case of Progressive Games involved a patent application for a modified version of a poker game which the court held not to be an “art” when applying the Shell Oil test as the game did not amount to “a contribution or addition to the cumulative wisdom on the subject of games” and “do not substantially modify the poker game as it exists nor do they create a new game”.

Progressive Games is another block facing the successful application of software patents as many of them are games or business methods. The new interpretation of Shell Oil in this case makes the satisfaction of subject matter issues a complex matter than requires examining novelty, non-obvious and utility at the same time. This cannot be right because it will give a patent examiner to have a “second kick at the can” for attacking the application based on these separate elements (A6). Some of the steps in the Shell Oil test also create a great level of uncertainty, as they are unlike the real separate stand alone elements, do not have established and clear tests for satisfying each of these steps (A6).

Regardless of the method for reaching the outcome of the case, it is clear from all previous court decisions that the court is not in favor of software and similar patents that do not involve a tangible element as it has never allowed them in all the instances that actually reached the court, even in the last decision of Progressive Games which was decided in 1999.

This dislike of software patents is not shared by CIPO. It will be seen from the upcoming post that CIPO has decided to use its own interpretation to allow some recent software patents to succeed even though they would have been rejected if the same court tests were applied.

Software Patents in Canada – The Position of the Courts (1)

Software Patents in Canada

The position of the software patents in Canada is not a straightforward one. There is no explicit prohibitions on the patentability of software patents (or business methods) in the Canadian Patent Act, but the only software patent case that ever reached the court decided that software inventions are not patentable. This case was decided in 1981 and the computer and internet revolution has changed the face of Earth since then and there seems to be a general preference for having patents for software in the Canadian Intellectual Property Office  – which indicates in its Manual of Patent Office Practice that a qualify software may be granted a patent. The CIPO did not stop there as it has already allowed the successful application of a number of software patents. CIPO decisions and MOPOP has no legal authority, so the fact that they are granting software patents does not mean that the court will hold them valid if ever challenged.

I will examine in this post the position of the courts and various grounds for which a software patent may be rejected and then will follow that with how CIPO has changed its position over time.

Requirements Under the Patent Act

For a patent application to succeed in Canada it must satisfy the invention subject matter requirements (Section 2), novelty (s28(2)), non-obviousness (s28(3)) and utility (s2).

For application to satisfy the subject matter requirements it must fall within the definition of “invention” in Section 2 of the Act which states that it must be an “art, process, machine or composition of matter, or any new and useful improvement in any art, process, machine, manufacture or composition of matter”. This definition is subject to sub-section 27(8) which prohibits the grant of a patent for “any mere scientific principle or abstract theorem”.

Novelty is a quantitative criteria of “newness” that requires the invention not to be part of the state of the art at the of filling.  (MOPOP S15.01.01) This basically means that the invention must not have been built, or described in a single document that contained sufficient information to allow another person to make the invention, before the application for the invention was filed.

Non-obviousness is a qualitative criteria of “inventiveness” that requires the invention not to have been obvious to a person of skill in the art at the time of filling. The test of obviousness is whether or not an unimaginative skilled technician would, in the light of the state of the art and common general knowledge at the time of filling date, be led directly and without difficulty to the invention covered by the claim. (MOPOP S15.01.02)

Utility is a requirement for the invention to be “useful” for the purpose for which it was designed. (MOPOP S15 + Progressive Games v COP)

Software patents are usually rejected on subject matter grounds and not on novelty, non-obviousness, or utility, which do not create any software specific problems to the patent application.

General Position of Software Patents – The Schlumberger Case

The only software patent case that reached the courts in Canada is the Schlumberger case (Schlumberger Canada Ltd. v. Commissioner of Patents (1981) 56 C.P.R. (2d) 204 (FCA)). The application in this case involved a computer system for the analysis of soil characteristics data for the purpose oil and gas exploration. CIPO rejected the application on the grounds that it was merely a computer program which did not fall under the definition of invention under section 2 of the Patent Act.

The application reached the Federal Court of Appeal which used a two step test to decide whether or not the application should be allowed:

  1. What has been discovered according to the patent application?
  2. Is that discovery patentable regardless of whether a computer is or should be used to implement discovery?

In applying this test the court held that “[w]has is new here is the discovery of the various calculations to be made and of the mathematical formulae to be used in making those calculations. If those calculations were not to be effected by computers but by men, the subject-matter of the application would clearly be mathematical formulae and a series of purely mental operations; as such, in my view, it would not be patentable,” and consequently rejected the application on the grounds of Section 28(3) against the patentability of a “mere scientific principle or abstract theorem”. 

This court decision made it clear that the mere addition of a computer in a patent application will not change the subject matter of the discovery so as to rescue what is originally unpatentable. A mathematical formula was not allowed to be patentable just because it is processed by a computer.

Though Schlumberger is the only case involving software patents and it was clearly not in favour of allowing software patents, the decision did not prohibit patenting ANY software related invention, but merely one in which “what was discovered” would not be considered as a qualifying subject matter if a computer is or should be used to implement it.

CIPO interpreted this a consequent decision (Application No. 96284) to mean the following:

  1. Claims to a computer program per se are not patentable;
  2. Claims o a new method of programming a computer are not patentable;
  3. Claims to a computer program in a novel manner where the novelty lies solely in the program or algorithm are not patentable;
  4. Claims to a computing apparatus programmed in a novel manner, where the patentable advance in the apparatus itself, are patentable; and
  5. Claims to a method or process carries out with a specific novel apparatus devised to implement newly discovered idea are patentable. 

This summary corresponds with the Schlumberger case as a pure computer program or an invention in which the novelty is in the program are not patentable, while an invention that uses a computer program and has its novel aspect outside the computer program would be patentable.

This position was justified at its time as the software industry was still in its early days and computer hardware was a bigger industry in which innovation was rapidly changing. Research and Development could have been easily protected by the patent system with the Schlumberger decision as the decision allows computer related inventions to be patentable where the innovation is in the apparatus (the hardware) and that adequately protects and promotes this industry.

However, this position is changed now as software is clearly one of the most important industries in world economy and there is a great need to protect and promote the investment in this field. This change in important in software is reflected in more recet CIPO decisions, which are not binding and can be overturned by any court if ever challenged.

But before we turn to the CIPO decision, we should talk in the second post about other more recent additional obstacles that face software patent application as became apparent from business method patent applications, which face a very similar objects at CIPO.