Discuss critically the current law and the proposed
reforms concerning the patentability of computer
programs
When the concept of protecting rights through Intellectual Property was in its infancy,
the IT industry was in its infancy. As the technological zeitgeist grows, the
issue of protection of rights related to computer programs grows with it. The
problem though is where to place the protection. Computer programs are written in a
mathematical type language which, in itself, would normally be protected by
copyright. Designers of computer programs would rather have protection of their
idea than the expression of it. The ideal way to protect ideas is by registering a
patent. That is not always easy though as the patent industry is hostile to computer programs.
This essay will attempt to address the current law and the premise of technical
effect. It will identify problems with the patent process and will analyse proposals
for reform.
Background
The European Convention on the grant of European Patents (European Patent
Convention) 1973 takes a hard line on computer programs by seeking to exclude
them from gaining patent protection. Section 52(2) concerns what shall not be
regarded as inventions (and therefore not being able to be patented). It includes-
�
(a) �mathematical methods
�(c)�programs for computers
(3) 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 the subject-matter or
activities as such.�1
Computer programs are clearly excluded as are the mathematical methods in which
they are written. The UK Patent Act 1977 in section 1(2) gives an almost identical
list of what cannot be classed as inventions. Again, �mathematical method�2 and a
�program for a computer�3 are listed. Importantly though, the same �as such�4 is
also to be found in the same section. This, as we shall see, has given the courts a
loophole in which to patent software.
Despite this, there are about 30,000 current patents on computer software in the EU
and about 15% of UK patents per year are now software based. The reason for this
contradiction lies in exceptions which the courts have found. Both the European
Patent Convention and the UK Patent Act were written at a time when the IT
industry was still in its infancy. Patenting computer software/programs was seen as
counter-productive as it could prevent the rapid improvement in IT. Though this
argument is still valid today, there is not the urgency there was when the legislation
was passed. As such, the courts are more willing to find exceptions for when
patenting computer programs is acceptable.
The Premise of technical effect
The major way this has been done is through the requirement of a �technical effect�
for patent applications not to be refused by section 1(2) of the Patents Act. This
requirement seizes on the stipulation that the application for a patent performs �as
such.� By making this requirement, computer programs can be patented provided
that the program is such that when run on a computer it provides a technical effect
which is more than would follow merely from the running of any program on a
computer, and which is such that claims to the computer when programmed would
not be rejected under section 1(2) under the existing practice.
This way round of patenting computer programs was found in the milestone case of
Vicom/Computer Related-invention 5 where the European Patent Office technical
board said-
�Generally speaking, an invention which would be patentable in accordance with
conventional patentability criteria should not be excluded from protection by
the mere fact that for its implementation modern technical means in the form of
a computer program are used. Decisive is what technical contribution the
invention as defined in the claim when considered as a whole makes to the known art�
This should not be taken to suggest that the courts have made it simple to patent
computer programs. Efforts have been made to patent computer programs without a
technical effect under different guises which have been rejected. For example it is
not possible to obtain a patent through providing a carrier with the program on. This
was the situation in Genentech Inc�s Patent where Dillon LJ said-
�It would be nonsense for the Act to forbid the patenting of a computer
program, and yet permit the patenting of a floppy disc containing the
program�it seems to me, that a patent�for the disk containing the program is
no more than a patent for the program as such�6
This is something though that the courts are likely to follow. The Technical Board of
Appeal of the European Patent Office has decided in two cases, both in the name of
International Business Machines Corporation7, that the practice of the European
Patent Office in under Article 52(2) and (3) of the European Patent Convention
should be changed in the respect of computer programs. The judgment approves the
decision that technical effect is needed for a patent to be granted and goes further
by recommending that the patent office practice should be changed in line with this.
Identifying the problems
Straight away, this presents a problem! How should technical effect be defined? The
only definitions given have been very vague, leaving uncertainty in the law. It is
accepted that a technical effect is generally an improvement in technology, and
needs to be in an area which is patentable. This has been criticized though as being
too vague and leaving plainly ridiculous results at times. For example, British
Telecom was able to patent the hyperlink which is a relatively short piece of
computer language. Whilst this simple script gained patent status, a very long script
which automates an accounts process without a technical effect would not. If
someone else independently created the same script for the accounts program, they
would escape copyright infringement and would have no patent to avoid. The patent
system is blind to the value of what is being patented.
The courts have also suggested that it is the �innovation�, not the implementation,
which matters. This further complicates matters as �innovation� again does not have
any clearly defined boundaries. It also leaves some worthwhile practices
discouraged. Developers of programs which automate processes or use other
complex programming script may not do so if they cannot protect their interests
through a patent. A similar situation was found in the case of Gales application 8,
which involved a claim to �Read only memory� (ROM) containing a program for a
computer. Mr Gale had put in a lot of effort and expertise in his creation, but could
not patent it. This left it protected only by copyright which left the path open for
others to create programs loosely based around his work. In rejecting the
application for patent, Nicholl LJ said-
�the attraction of Mr Gale�s case lies in the simple approach that�he has found
an improved means of carrying out an everyday function of somputers�A
computer�will be a better computer when programmed with Mr Gale�s
instructions�but the instructions do not embody a technical process which
exists outside the computer. Nor, as I understand the case as presented to us, do
the instructions solve a �technical� problem lying within the computer.�
As �technical effect� is so ambiguous it is not clear what can and cannot be
patented. With so many new computer program patents each year, the patent system
has become unpractical for software developers. With approximately 30,000 EU
software patents in existence, it is not practical for software developers to trawl
through each of them checking if they are writing something which has already been
patented. Most inventions may have one or two patents on them. As such, it is
relatively easy to check if any patents would be infringed. With computer programs
though containing thousands of lines of script, there may be dozens, if not hundreds
of patents for each program.
Further to that, because the patent system is so slow, it can often take about 18
months for a program to be patented from application. During that time information
on the patent is classed as �patent pending�. The problem that presents is that
software can be written and even released whilst a patent is still pending. This is a
very common occurrence as software developers will often create new programs in
response to new technology or a new market, at exactly the same time and without
knowledge of each other. If a program is released whilst a patent is pending, the
holder of that patent is still able to sue for breach. There is no fault present and yet
a claim for breach is still valid. If the computer program were protected by
copyright instead, there would be no claim. In copyright, material can be identical
IF it was written in ignorance of the other persons work.
It must also be considered that not all of the people who are affected by software
patents have the same wishes. Relatively small scale enterprises, one-man bands and
part timers are unlikely to have the time to search through long lists of patents to
ensure they are not infringing any. They may even be discouraged by the cost of
patenting their own work or by the likely damages that could be awarded against
them if they were to infringe someone else�s patent.
Large companies would have more time to invest in checking lists of existing
patents, but could also lose large amounts of money invested in projects if someone
else already has a patent over part or all of their work. Normally, these large
companies would be protected by insurance against such claims, whilst small and
medium sized enterprises (SME�s) would not. The reality this leaves is that SME�s
can be discouraged from innovation and insurance companies can charge high
premiums to large companies.
Proposals for reform
"Computer software as such is excluded from patentability under the European
Patent Convention and the laws of the Member States, and the Commission has
no intention of changing this situation."9
This was the answer given by Mr Bolkestein on behalf of the European Commission
in response to a written question on the future of computer software patents. It is
not strictly the case though. By avoiding the essence of the question, the
Commission avoided giving an answer they are not sure of. Mr Bolkestein answered
the question on programs on such.
In the United States, computer programs can be patented regardless of whether they
have a technical effect. That has lead to a barrier to entry for SME�s and a growing
number of claims for patent breach against large companies which control the
software market. Though few people in the UK or the European Union Institutions
are proposing anything as drastic, the European Commission proposals would extend
the scope for computer program patentability.
The proposed directive circumvents article 52 of the European Patent Convention
through application of some principals found in the case law to it. The Commission
aims to introduce the concepts of �technical effect of software� and �technical
contribution of software as ways in which software can be legitimately patented.
Though the commission has stated that for software to be patented, it must make a
technical contribution to the state of the art. As this does not refer to any real
physical or measurable effect of the software, the terms used still give some
uncertainty. In essence, the commission�s proposals seek to give case law principles
codification in the Convention. If done so, the UK would also have to comply.
Along with the lack of clarity, the major problem the proposals leave is that SME�s
are still left at a disadvantage. As more and more patents for computer software are
granted-barriers to entry to the market grow. It may be difficult to find out
which existing patents may affect the program under development. There may
be thousands of patents to check. Patents pending will remain secret and impossible to
discover details of, leaving a lottery of whether software developer�s work already
has a patent pending. When Patents are granted, they are written in such a precise
legal style that the average software developer may need expensive legal help just
to check whether their activity will breach a patent. As it is often discovered at a late
stage that a patent exists over a specific algorithm or feature, a license may need
to be bought. As this can often be expensive, a common practice is for a trade-off of licenses.
This generally only benefits large company�s who have numerous patents. Overturning a patent
in court can also be very expensive.
Conclusion
If patents are designed at encouraging innovation, they must do so by exercising the correct
balance between protecting the rights of patent holders and ensuring software designers
do not have to tread so lightly, as to hinder their efforts. Many programs �as such� are
innovative and yet cannot currently be patented. The Patent process for computer programs
has become a time consuming lottery. Software designers write computer programs in the hope
that they are not infringing the vast amounts of existing and pending patents without knowing
what they all are. They then
patent as much of their software as they can in the hope that it will make them a lot of
money. In reality, the vast amounts of patents, like lottery tickets do little but cost
money and amount to something worthless. Only a minority of software patents actually
give the owner a genuine benefit.
The proposals for reform do little but codify the existing case law. Whilst the government
calls for an increase in the number of patents on computer programs, software developers
in SME�s are opposing nearly everything that is said. Opinion in large companies is divided
between those wanting to protect their large investments and those wanting to avoid large
insurance premiums needed for the event of breach being claimed by another. To encourage
competition in the IT industry, it is essential that SME�s are not excluded through financial
and practical barriers from operating in the market.
Whilst it is reasonable to accept that computer programs should be patented where there is a
technical effect, there does need to be much more of a limitation. If fees were hiked
to patent software, companies would be deterred from patenting everything but the essential.
These fees could be reduced for those at the lower end of the market in order not to exclude
them. With the patent office gaining higher revenue and lower numbers of applications, the
time from application to patent would be reduced. This would ease the problem of patents
pending. Where claims are currently made for breach of patents which were pending at the time
of software release, licenses could be granted on a limited basis only to those breaching
the patent (for example, that the patent holder can only distribute software containing
the patent to a fraction of the patent holder).
Whilst a requirement for technical effect is reasonable, stipulations of value to the
patent should also be considered when deciding whether to grant the patent.
By doing the above, the patent system would become less of a lottery and more practical. It
would encourage innovation and would not discriminate against SME�s. Without these changes,
or if the Commissions proposals are made law, numbers of patents will increase and the
software system will overload. There certainly need to be changes to clarify the law and
protect genuine interests. Put simply-
�In cases when it (the bureaucracy of the patent system) doesn't serve any purpose
except to create artificial monopolies so that somebody can interfere with software
development, squeeze money out of developers and users, then we should reject it.�10
A reduction of the number of patents would be beneficial to innovation, not an increase.
The Commissions proposals are a �fatal error�.
WORDS 2493
Citations
1. European Convention on the grant of European Patents s52 (2) and (3)
2. Patent Act 1977 s1(2)(a)
3. ibid
4. ibid
5. Vicom/Computer related invention T208/84 OJEPO 1/87
6. Genetech Inc�s Patent (1989) RPC 147 Dillon LJ at page 240
7. International Business Machines Corporation (2 cases) T0935/97 and T1173/97
8. Gales Application (1991) RPC 305
9. Answer given to Parliamentary written question E-3927/02 by Mr Boklenstein on behalf
of the Commission, 11 February 2003
10. Stallman, Richard. Speech - "The danger of software patents":
http://lpf.ai.mit.edu/Patents/danger-of-software-patents.txt
Bibliography
1. Intelectual Property: Patents, Copyright, Trade Marks and Allied Rights by Cornish
& Llewellyn, fifth edition
2. Cases and Materials on Intellectual property by William Cornish, fourth edition
3. Stallman, Richard. Speech - "The danger of software patents":
http://lpf.ai.mit.edu/Patents/danger-of-software-patents.txt
4. Bricklin, Dan. (Dan Bricklin invented the electronic spreadsheet but did not patent it):
"Patents and software": http://www.bricklin.com/patentsandsoftware.htm.
5. Lale, Chris. "Software patents- a musical analogy":
http://www.coolscience.co.uk/free-software/software-patent-music.html.
6. Lale, Chris. "Software patents- the technical
effect":http://www.coolscience.co.uk/free-software/software-patent-techeffect.html