Reading the "Proposal
for a Directive on the patentability of computer-implemented
inventions - frequently asked questions" raises many more
questions that it answers, this document provides a commentary. |
First to clarify what is a patent? There's two
perspectives:
As a result patents do not simply provide ownership of ideas for a limited duration for a company as intellectual property proponents wants to claim. Patents are a trade between a company and the society. The society buys an idea from the company, but pays for that idea during the 20 years period where the company is free to exploit the patent monopoly society has granted it. |
| Legislation: The following articles from the European
Patent Convension (EPC) defines patentable subject matter, e.g.
which ideas that can be patented:
EPC Article 52 Patentable inventions (1) European patents shall be granted for any inventions which are susceptible of industrial application, which are new and which involve an inventive step. (2) The following in particular shall not be regarded as inventions within the meaning of paragraph 1:
The provisions of paragraph 2 shall exclude patentability of the subject-matter or activities referred to in that provision only to the extent to which a European patent application or European patent relates to such subject-matter or activities as such. EPC Article 54 Novelty (1) An invention shall be considered to be new if it does not form part of the state of the art. (2) The state of the art shall be held to comprise everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the European patent application. [..] EPC Article 56 Inventive step An invention shall be considered as involving an inventive step if, having regard to the state of the art, it is not obvious to a person skilled in the art. [...] |
| Practice: The following is sections from the EPO
guidelines on computer programs:
.
Thus in the 23 years since 1978 the EPO interpretation of the EPC has been essentially reversed with regards to software patentability. This is also seen in the some 30000 software patents issued by the EPO before 2001. National courts have to a large degree followed the original guidelines of 1978, which were formulated shortly after the European Patent Convension itself was formulated, and thus it's likely that the EPO guidelines followed the original interpretation of the EPC. |
What are the main features of the proposal?
The central requirement of the proposal is that, in order to be patentable, an invention that is implemented through the execution of software on a computer or similar apparatus has to make a contribution in a technical field that is not obvious to a person of normal skill in that field. This is essentially a legal question of a kind which is answered all the time by patent offices and practitioners.
| So what is new in this proposal?
EPC article 56 already states that an idea can be patented if it provides a contribution to the state of the art which is not deemed obvious to a person skilled in that art. While the EPC only requires industrial application, only technical features of an idea can be be protected by a patent as it is clear from the following EPC Rule 29 Form and content of claims (1) The claims shall define the matter for which protection is sought in terms of the technical features of the invention. [..] The claims are the part of a patent which describes what applications of the patented idea that a monopoly is applied for. So why do the EU issue a directive and a press release where the novelty is that it was already included in a convension from 1978? The directive requires that the EU member states modifies their patent legislation such that software can be patented, irregardless of whether or not that software is bring used to control a industrial applicable process which manipulates natural forces to produce some product, e.g. a technological patentable invention. |
Does this proposal follow the principles established by the European Patent Convention?
Broadly, yes. The European Patent Convention says that patents should not be granted for computer programs as such. However computers are themselves machines like other technological devices. Over the years, national courts have decided that there is no reason why a patent should not be granted for a machine programmed to carry out some technical function, or a technical process carried out using a computer or similar machine. But in common with all other inventions, it still needs to be new and not obvious. The proposed Directive has followed this approach.
| Broadly? Compare the EPO guidelines from 1978 and 2001
both of which should follow the principles established by the EPC,
and they certainly seems to cover a broad spectrum of interpretations.
According to the original EPO guidelines formulated shortly after the formulation of the EPC itself, it's made clear that software is not patentable, however, since then the EPO has allowed software patents, which has put the national courts under pressure to adhere to the EPO interpretation of the EPC. Notice also that even the EPO 2001 guidelines mentions a EPO appeal board decision that software does not become technical just because it runs on a computer which is a technological device, as it is argued above. The T 26/86 Koch & Sterzel/X-ray apparatus (see also T1173/97) EPO appealboard decision found that a program generated electrical currents in the computer (which is technological device), but because the currents only represented information they could not be construed as a technical effect of the program function. The proposed directive goes even beyond the very weak technicality requirements of the EPO. Patent granted for at machine programmed to carry out a technical function (such a machine that cures rubber which is controlled by a computer) does not infer that the software controlling that machine becomes technical, just as the brain of a human 'programmed' (e.g. paid and trained) to carry out the same technical function does not become technical and thus patentable subject matter. Nor would reading a book by a technological device, such as for instance a scanner-optical character recognition-speech syntehesis device make the content of the book technical, even though the content of the book would be instructions for the technological device on what to say out loud (the technical effect of the book when executed on a computerized book-reading device). |
What is meant by "technical contribution"?
The proposed Directive defines a "technical contribution" as "a contribution to the state of the art in a technical field which is not obvious to a person skilled in the art". If an invention implemented through the execution of software did not make a technical contribution it would be considered to lack an inventive step and thus would not be patentable.
| Software is technical because it runs on a technical
machine (a computer). Thus 'technical contribution' means that the
patentable idea has to provide a contribution, that can run on a
computer, which is not obvious to somebody skilled in programs that
runs on computers. If an invention implemented as a program can not
run on a computer is lacks an inventive step and is not patentable.
This is also clear from the directive itself:
Short: National courts shall recognize anything that runs runs on a computer is technology, thus patentable if it's non-obvious. |
It would not be possible for a legal text such as a Directive to attempt to spell out in fine detail what is meant by "technical", because the very nature of the patent system is to protect what is novel, and therefore not previously known. In practice the courts will determine in individual cases what is or is not encompassed within the definition. However, earlier court decisions have indicated that a technical contribution may arise if there has been some improvement in the way that processes are carried out or resources used in a computer (for example an increase in the efficiency of a physical process), or if the exercise of technical skills beyond "mere" programming has been necessary to arrive at the invention.
In other words the it's not possible to spell out
sufficiently clear details about what is patentable, as to clear up
the any debate about the limits of patentable subject matter, or the
application of such limits. The german federal court have made a
definition of technical: "using natural forces to produce a causally
expected result without the intervention of human intelligence".
However, this definition rules out patents on ideas in software, so
it must be wrong..
This is not consistent! The second example is what EPO calls a 'further technical effect', software that increases security, or decreases resource requirement such as harddisk space, CPU time, or time to send data over the network are given as example of 'further technical effect'. Algorithms (which the PR excludes from patentability) can have a 'further technical effect' (and thus be patentable). An example is compression algorithms. Compression algorighms concern themselves solely with the naure of data (data in-data out algorithms independent of any semantical meaning of the data in terms of sound, pictures, text, or numbers relating to properties of a physical entity), however, since compression algorithms reduce the harddisk space required to save the data, such algorithms have 'further technical effect', even though they do not have a technical effect because they just transform data, into another data that contains the same information but requires less space. Reading the directive proposal it's not apparent that algorithms solely concerned with the nature of data or how an application operates on data is extempt from patentability. Probably this statement in the Press release relies on a subjective interpretation on technical, not supported by the directive proposal itself. The directive requires that the national courts accept the concept of 'further technical effect' as a being technical effect, even though it's just data processing. |
What is the current situation in the EU concerning patents on computer-implemented inventions?
Does it make sense to call software an invention? |
Inventors currently have two possible routes for obtaining protection for their inventions. Patents may be applied for, processed and granted either at the European Patent Office (EPO) under the centralised system of the European Patent Convention (EPC), or via national patent offices in the Member States purely according to national law. However whichever route is chosen, national law applies in all cases after grant. Thus, granted European Patents become a "bundle" of national patents which have to be validated, maintained and litigated separately in each Member State.
| Which assumes that it's possible to get a software
patent from a national patent office. The EPO has issued 30000
software patentes, how many software patents have so far been issued
by national patent offices?
According to the explanation of the directive software patents are not available from all European Patent Offices, probably they are only available from a few patent offices besides the EPO. |
Legally speaking, the patent laws of the Member States are in principle supposed to be consistent with the EPC and uniform among themselves. However, in practice there is no unifying structure with binding effect on national courts and there is therefore the potential for differences to appear over the interpretation of particular aspects of patent law. The EPC is entirely separate from the Community and the EPO is not subject to Community law. The EPC currently includes among its membership all Community Member States plus five other countries (Switzerland, Turkey, Cyprus, Monaco, and Liechtenstein). Several more are expected to join later in 2002.
The patentability of software and related inventions is currently determined principally by Article 52 paragraphs (2)(c) and (3) of the EPC, according to which computer programs "as such" (as well as business methods and certain other entities) are excluded from patentability. However, since the EPC came into force in 1978, more than 30,000 software-related patents have been granted, and a considerable body of case law on the subject has been built up by the appellate bodies of the European Patent Office (EPO) and the Member States’ courts. Many of these patents have been granted for devices and processes in technical areas, but the majority now relate to digital data processing, data recognition, representation and information handling. This has fuelled debate on whether the limits of what is patentable are still sufficiently clear and applied properly.
Should modifications of patentlaw be determined by
case law or by political decisions? It's the society that pays for
software patent monopolies, and the politicians are the democratic
representatives of the members of societies, who do the appeal board
of the EPO represent, who elects EPO appeal board members? |
This is moreover an area in which some differences have appeared in practice under national patent laws. This has led to uncertainty over the legal position especially in Member States where less extensive litigation has taken place.
| Is there any national patent laws that clearly allows software to be patented as is suggested by this directive? In how many member states has extensive litigation taken place? And has these rulings in these few states been clear and consistent? |
This is therefore an opportune time to propose EU legislation. In its absence, Member States may feel obliged to act outside the Community framework. In addition, if no action is taken at the Community level, the scope of what is patentable may be defined on the basis of individual decisions on particular cases by the European Patent Office’s (EPO) judicial bodies, without the opportunity for coherent political reflection based on wide consultation and the overall picture. The proposed Directive therefore sets clear borders to what would be patentable in the EU and what would not.
What EU citizens has elected the members of the EPO
judicial bodies represent? What democratic control does the EU have
over the EPO practice and it's juridical bodies? None! |
Why are patents good for innovation?
| Begging the question: Are patents always good for innovation? |
From the perspective of the owner of a patent, the limited exclusivity provides an opportunity to generate income. This in turn provides an incentive to invest in research and development. And a successful business will use income from patented inventions to fund further innovation. Patents can also be used as security to obtain loans and as a means of negotiating for licenses on technology owned by others.
It is a condition of obtaining a patent that there should be disclosure of how the invention works. If the disclosure is not sufficient to enable the invention to be reproduced, the patent can be revoked. Patents are therefore an important source of technical information for others to use.
EU industry has built itself up in the very legal environment which we are now seeking to harmonise. Other countries which are successful in the information technology sector (such as the US and Japan) also grant patents for computer-implemented inventions.
| What about from the perspective of the society who will
pay for the creation of patent monopolies? Is it a good idea to allow
any program to be patented? Since the majority of software patents
are trivial if a good prior-art search was performed, thus the value
of IPR is just the next IT bubble, which will burst at some time.
Remember a bubble only exists as long as everybody believes it's real.
Regarding licensing does it make sense to parsel out ownership of software ideas, because others already has ownership of ideas and wants licence fees? Are the information sector successfull in the US and Japan because they grant softwarepatents, or inspite of software patents? |
What is the difference between a patent and copyright? Can an invention be protected by both?
Patent and copyright protection are complementary. In very general terms, patents protect new technical ideas and principles, while copyright protects the form of expression used. For example, a new sort of paper might be protected by a patent, while the printed content of a newspaper would be protected by copyright. In computer terms, the actual code (whether machine-readable or in a form which is intelligible to human readers) would be subject to copyright protection, while underlying technological ideas may be eligible for patent protection.
Patent law gives the holder of a patent for a computer-implemented invention the right to prevent third parties from using software incorporating any new technology he has invented (as defined by the patent claims).
On the other hand, a computer program will be accorded copyright protection where the form of expression is original in the sense of being the author’s own intellectual creation. Third parties would not be able to produce substantially the same content material as the original author has produced, even if they used different technical principles to do so.
Accordingly, the same program may be protected by both patent and by copyright law. That protection may be cumulative only in the sense that an act involving exploitation of a particular program may infringe both the copyright in the code and a patent whose claims cover the underlying ideas and principles of the invention using the program.
Isnt software different to other technologies in that patents can be used to block legitimate independent innovation?
The Commission has seen little evidence that this has been a problem in practice in the present environment. This would be the case only if the scope of protection granted by patents were extended to software as such and, for instance, blocked the use of an algorithmic idea in other technical fields from the one in which a patent is granted. Since the proposed Directive would not extend the scope of what can be patented, nor the scope of the protection granted by a patent, there should be nothing to fear on this front.
| Have the Commission looked for such evidence? E.g. in
the 90% of the hearing replies from the Eurolinux alliance?
What does "patents on software as such" (which is currently
covered by copyright) have to do with the question that patents
blocks independet innovation because software is different from
technology? Isn't software different is asked, today if somebody invented a
novel way of using the wheel in a technological application, they
could get a patent on the wheel, but that patent would be limited to
the scope of the technological field as defined by the novel
application of the wheel. The difference between technological
inventions and software is that a software patent limits all
applications of the patented idea that runs on a computer. |
Would the proposed Directive restrict the interoperability of computer programs?
No. In fact, Member States’ patent laws, while not fully harmonised, do not in general extend to acts performed privately and for non-commercial purposes, or to acts carried out for experimental purposes related to the subject-matter of the invention.
| Example: If Microsoft has a software patent that is
required to reading or writing Word files, then people may still
create programs that reads and writes Word files, as long as these
programs are NOT used commercially, or only for experimental purposes
of reading or writing Word files. If a competing company wants to
create, use, sell, import, or distribute software that can read or
write Word files, then that competing company is required to get a
licence from Microsofts. That restricts interoperability.
(Note this is an example, and I do not know if Microsoft has patented the Word format, or algorithms required to use it, however replace word files by microsoft .ASF multimedia files, and they have a patent, which has been used to stop independent development of a program that could display these files.) |
Nor is it likely that the making of a back-up copy in the context of the authorised exploitation of a patent covering a programmed computer or the execution of a program could be construed as an infringement. The proposed Directive will not change this situation. Thus, because of the differences between the subject-matter of protection under patent and copyright law, and the nature of the permitted exceptions, the exercise of a patent covering a computer-implemented invention should not interfere with the freedoms granted under copyright law to software developers by the provisions of the existing Directive on the legal protection of computer programs (91/250/EEC).
Indeed, the proposed Directive makes specific reference, inter alia, to the provisions on decompilation and interoperability in the Directive 91/250/EEC, in terms that ensure that the different scope of protection granted by patents would not undermine the possibility to carry out the acts that are permitted under that existing Directive. Directive 91/250/EEC includes specific provisions (Articles 5 and 6) to the effect that copyright in a computer program is not infringed by acts which under certain circumstances would otherwise constitute infringement. These exceptions include acts performed for the purposes of studying the ideas and principles underlying a program and the reproduction or translation of code if necessary for the achievement of the interoperability of an independently-created computer program. It is also specified that the making of a back-up copy by a lawful user cannot be prevented. Such provisions are necessary in the context of copyright law because copyright confers the absolute right to prevent the making of copies of a protected work. All the acts mentioned involve making copies and would therefore infringe in the absence of any exception.
What are the main differences between the approach of the proposed Directive and the situations in the US and Japan?
The main difference lies in the requirement for "technical contribution". Japanese law does not have this as such, but there is in Japan a doctrine which has traditionally been interpreted in a similar way: the invention has to be a "highly advanced creation of technical ideas by which a law of nature is utilised".
In the US on the other hand, a patentable invention must simply be within the technological arts. No specific technological contribution is needed. The mere fact that an invention uses a computer or software makes it become part of the technological arts if it also provides a "useful, concrete and tangible result". Among other things, this has meant that in practice in the US, restrictions on patenting of business methods (apart from the requirements of novelty and inventive step) are negligible.
Practice in both the U.S. and Japan is to allow patent claims to software that implements patentable inventions.
The need to maintain and improve standards of patent examination for computer-implemented inventions is one of the most important challenges facing patent offices world-wide. The EPO, which is widely recognised as a world leader in patent search and examination, is, together with its colleagues in the US and Japanese patent offices, studying this question as one of its highest priorities.
As argued about the directive criterion for technical patentable subject matter is if it runs on a computer. Hence, it's similar to the US doctrine. Claiming that the EPO is a world leader in patent search and examination is a really scary thought, given the horror gallery of software patents it has issued. (REF). |
Wont extending patents to cover computer programs reduce competition?
The Directive will not make it possible to patent computer programs "as such". In broad terms, nothing will be made patentable which is not already patentable. The objective is simply to clarify the law and to resolve some inconsistencies in approach in national laws.
The answer does not seem to be
related to the question. |
Does the proposal include rules about business methods?
The proposal does not deal directly with the patentability of business methods. This is because the consultations have indicated clearly that there is general satisfaction with the current situation, whereby "pure" business methods (that is, methods which have no technical character) are not currently patentable.
However, in practice some inventions involving business methods could fall within the definition of "computer-implemented inventions". These inventions would be dealt with in accordance with the proposed Directive, and in particular patents would only be granted for inventions that made a "technical contribution".
| Translation: business methods that can not run on a computer is still not patentable (people are still free to buy food on the market). However, if a business method can be run on a computer, and is not obvious to somebody skilled in the art of running programs on computers, then they are patentable. The directive makes it clear that the novelty value does not need to reside in the technical contribution, hence novel business metods that runs on computers such as e-commerce can be patented. |
Why would claims to computer programs on their own not be permitted under the proposed Directive?
In recent decisions, patents have under certain conditions been allowed which contain claims for computer programs on their own, for example on a disk or even as a signal transmitted over the internet. In the course of consultations, fears were expressed that if enforced, patents including such claims may be used to prevent "reverse engineering" and other activities considered legitimate in respect of computer programs already protected under copyright law. Moreover, such claims could be said to be contrary to the EPC, which does not allow patents for computer programs "as such". In response to these concerns, the Commission has decided not to follow the direction taken by case law in this important respect. Accordingly, the proposal would not allow claims of this type to be considered valid.
Just to clarify, selling, producing, importing,
distributing and running programs for commercial purposes can
infringe patents. Reverse engineering is still allowed (in
copyright), but commercial usage of a program that uses patented
ideas for instance obtained by reverse engineering still require a
licence from the patent owner. |
What about the BT "hyper-text" patent case - can this patent really be valid when hyperlinks are common-place?
This is a very old patent. It has expired everywhere else in the world, but is still in force in the US by virtue of the old US patent law which counted the patent term from the date of grant (10 October 1989) rather than from the date of application. This happens from time to time and is nothing to do with being in the field of computer programs. It of course remains to be seen whether the US courts will find in favour of the patentees.
A related patent was granted in the UK and remained in force until it expired in 1997. Patents were also granted in many other countries including Japan and the main European countries. The test for whether a patent is allowable is whether it was inventive at the time of application. Although hyperlinks are commonplace today, it is clear that several different patent offices all took the view that the patent described a valid invention at the time of the application.
| The BT hypertext patent has prior-art, hence that is was granted by several patent offices just goes to show the fallacy of prior-art seaches. |
Would the Amazon "one-click" shopping cart ordering model be patentable under the Commission proposal?
The European Patent Office has yet to come to a decision on the related European application, so it would not be appropriate to comment on whether there is any patentable subject-matter in the application as a whole. However, a patent with the breadth of claims which has been granted in the United States would be highly unlikely to be considered to make a "technical contribution" in the EU under the terms of the proposed Directive.
| The EPO has issued the Bellboy patent (EP738446), which is a patent on services offered via e.g. websites, it has also issued a patent on paying bills via online banking (EP504287), and online auctions (EP900424). Where is the novelty value in paying bills? |
So why do we need a Directive?
The present legal situation has essentially been built up by national courts and the EPO. They have had to develop interpretations to cope with a technology whose subsequent development could not have been imagined when the European Patent Convention was first drafted at the beginning of the 1970s. They have done a very good job, but the lack of harmonisation and the existence of different legal traditions has meant that some differences have arisen with the potential for more serious divergences in future if action at the Community level were not taken. Consultations undertaken by the Commission on this issue have indicated very clearly that the lack of legal certainty in this field is widely regarded as very damaging to European interests.
So the definition of technology has to be adapted to
what companies want to patent (from the press release: Current law on
this question was drafted in the early 1970s when there was no
inkling of what was to come in the shape of modern computers and
networks, not to mention the emergence of a software industry worth
billions of euros.'). Thus it's probably not to mention what if it's
in the the interests of the society to grant software patents. |
What consultations has the Commission undertaken on this issue?
There has been a series of consultations on this subject beginning with the 1997 Green Paper on the Community Patent. Most recently, a discussion document was published on the Europa website on 19 October 2000 (see http://europa.eu.int/comm/internal_market/en/indprop/softpaten.htm, which invited comments by 15 December 2000 on the basis of a number of proposed "Key Elements" for a harmonised approach to the patentability of computer-implemented inventions in the European Community. These "Key Elements" broadly reflected the present state of development of European case law under which it is estimated that at least 30,000 patents for computer-implemented inventions have already been issued.
The consultation produced 1447 responses, the overwhelming majority by e-mail. An analysis of these responses is available at:
http://europa.eu.int/comm/internal_market/en/indprop/softpatanalyse.htm
| Commentary can be found here. (REF) |
Many of the responses supporting a more restrictive approach than at present, with fewer patents being granted, were transmitted through an open forum set up by the "Eurolinux Alliance", a group of companies and other entities supporting the development of open source software such as Linux. Although this group numerically dominated (90%) the response, the major sectoral bodies representing the information and communication technology industries, as well as many of the Member States, all supported the approach put forward by the discussion paper. Some responses argued for eligibility for patents to be widened in line with the practice in the US.
A democracy usually works based on numerical majority. But since hearing answers are due to self-selection it does not make sense to make statistics over the replies or to argue about who has a majority. However, who are the most innovative companies, the few large with huge IPR interests and -lawyers, or the many small companies that are busy innovating, and has no time for hearings on software patent issues? |
How does the proposal reflect responses to consultation?
The proposal takes account of all responses, which covered a huge spectrum of opinion and interests. The objective is to achieve the right balance between making patents available where appropriate in order to reward and encourage innovation, while avoiding stifling competition and open source development. The Commission received some submissions arguing that patents tend to restrict innovation in fields like software development. We also received submissions from organisations representing many thousands of companies arguing that computer-implemented inventions should remain patentable or even that patentability should be extended. The Commissions proposal reflects the balanced interests of the EU’s economy and society as a whole.
In what sense does allowing patents on anything that
runs on a computer and is not obvious to somebody who knows about
programs that runs on a computer take into account the anti-software
patent responses? It's difficult to imagine a more pro-software
patent directive. |
When would the proposal become law?
Once the proposal has been adopted by the EU’s Council of Ministers and the European Parliament under the so-called co-decision procedure, the Directive would have to be implemented in national law by the Member States.
Will there be further consultations?
There have already been extensive consultations since 1997 and no more are
planned now that the proposal, reflecting the results of these consultations,
has been presented. However, there is plenty of opportunity for further input
and comment on the proposal during the process of its adoption by the EU’s
Council of Ministers and the European Parliament.