For the half-year to 31 December 2014, the IPKat's regular team is supplemented by contributions from guest bloggers Rebecca Gulbul, Lucas Michels and Marie-Andrée Weiss.

Regular round-ups of the previous week's blogposts are kindly compiled by Alberto Bellan.

Friday, 25 January 2008

High Court allows computer program patent claims


In a surprising (to this Kat at least) turn of events, the Honourable Mr Justice Kitchin has ruled today that the current UK Patent Office practice of flatly rejecting patent claims to computer program products is wrong. The judgment, in the matter of several GB patent applications in the name of Astron Clinica and others, is now available from BAILII here. Many thanks go to John Gray of Murgitroyd & Co. for originally passing the news on.

After comprehensively summarising the last couple of decades of legal developments, covering the usual suspects (Gale, the oft-misspelled Merrill Lynch, Fujitsu and various EPO decisions), Kitchin J arrived at the main question in this appeal, which was whether the UK-IPO was correct in construing that the Court of Appeal judgment in Aerotel/Macrossan inevitably prohibited the patenting of all computer programs, or whether the old approach of considering the 'potential' technical effect of a computer program (following the EPO approach) could be taken into account, in a similar way to considering the effect of a method claim that would inevitably be carried out by running a program (which all of the applications under appeal contained). The UK-IPO had concluded that Aerotel/Macrossan ruled out computer program product patent claims, and consequently reverted to its old practice of rejecting such claims.

Kitchin J, however, considered that the point did not actually arise in Aerotel/Macrossan, because the court was not even asked to consider the question of computer program products claims. Although the Court of Appeal had criticised many EPO decisions, it had not criticised the main decisions relating to this point, being T 1173/97 and T 935/97. Also, the new four step test approach should produce the same result as the 'old' approach, and the Court of Appeal had said as much by saying that Merrill Lynch must be followed.

Probably more importantly, Kitchin J recognised that it was highly undesirable to have provisions of the EPC construed differently at the EPO as compared with the courts in the different contracting states, and that decisions of the Boards of Appeal should be highly persuasive. Mention was also made of the contrasting approach taken in Germany, where the EPO line tends to be followed closely.

The apparent approval of the UK-IPO's rejection of computer program product claims in Oneida Indian Nation's Application [see IPKat commentary here] was either rejected by Kitchin J as not actually meaning that, or was in the alternative respectfully disagreed with (as the High Court is allowed to do, in contrast with the Court of Appeal), depending on the different possible interpretations of Christopher Floyd's judgment.

In conclusion then, Kitchin J found that the appeals should be allowed. Each application concerned a computer related invention where the examiner had allowed claims to, in effect, a method performed by running a suitably programmed computer and to a computer programmed to carry out the method. The Hearing Officer had rejected corresponding program claims on the basis that they were necessarily prohibited by Article 52, and in Kitchin J's judgment he had erred in doing so. The cases were remitted to the UK-IPO for further consideration in light of the judgment.

The IPKat is, frankly, quite amazed by the judgment, because he was (apparently quite wrongly) convinced that there was no room for manoeuver after Aerotel/Macrossan, in particular in light of one of the central points of A/M being that the scope of the monopoly must be considered when construing the claim, which appears to have formed the basis of the UK-IPO's change of practice. He wonders whether the story has run its course for now, and if we can simply all go back to falling into line with the EPO, or if the UK-IPO will judge that this one is worth going further on. Will they? Can they? Should they? What would/does the embattled Lord Justice Jacob think? The IPKat would very much like to know, and he suspects his readers may be just about to tell him...

Catch up with the issues from the IPKat's previous posts here, here, here, here, here, here, here, here, here and here (to cite just a few).

28 Jan Update: This post has now been linked to, and commented on, in (at the latest count) Slashdot, I/P Updates, Patently-O, the IAM blog, and IP:Jur, as well as in Greg Aharonian's latest Patnews email. Many thanks to all. Expect something to appear shortly on Out-Law and the Register.

27 comments:

Cristian said...
This comment has been removed by the author.
Gobhicks said...

Frankly, Gobhicks is surprised that the Kat is so surprised.

The "program claim" issue has clearly been quite distinct from the "technical matter" issue, in UK and EP.

If a method claim is acceptable, and the technical inventiveness in the method claim is expressed commercially in a software offering, why should the software product that enables the technical invention not be directly protected by an appropriate patent claim?

This not the thin end of a wedge, it is the only logical conclusion based on established UK/EP reasoning.

Simon said...

I think that most of the concern about software patenting stems from a fear that we would end up following the US down the route of patents such as the Amazon 'One-Click'. It seems to me though that the main issue in the US is the USPO's very generous (other people might be less polite) approach to inventiveness and prior art. I would expect the EPO or UKIPO to be far more stringent.

I am, nonetheless, still uncomfortable with the idea of software patents because I feel they represent clear blue water between business process patents (which are an abomination) and 'conventional' technical patents, which are quite legitimate. Even if this case stands, and it just moves the boundary somewhat, it will lead to patents being considered for programs that are (cf Macrossan) just business processes in disguise.

Finally, I can't see this case not going to appeal. At the very least, the UKIPO will want the Court of Appeal to review its decision in Aerotel/Macrossan; we may even see the issue be remitted to the Lords for a final (?) decision.

Filemot said...

Congratulations are due to a fine victory for a Patent Agent Litigator against a shiny new QC. As I read it the claims are not allowed just back with the examiner to apply the four step test and get the same answer as for the method claims. Its simple, the whole of each application now gets rejected by the UK IPO.

Anonymous said...

I think Cristian has possibly got hold of the wrong end of the stick here - possibly Simon as well. This case was not about whether the companies could get patents for their inventions - the Office was prepared to allow various claims in each application. The question was whether they could have claims to just the computrer program (which is probably what they actually wanted to sell) or whether apparatus claims had to also include the computer that was going to run the program.

To filemot - as I understand it the examiners have already applied the four step test and some of the claims have passed. I can't see the applications being rejected now.

To Simon - I'm not sure the Office would appeal - it is rare for the Office to appeal if it loses. One exception was in Gale, where the High Court judgement essentially said that any claim to a ROM containing a program was patentable.

AngryTechie said...

Is there any way for third parties negatively affected (i.e. real software developers) by the introduction of corrupt EPO-style software-patents-in-disguise to challenge it?

I'm not a lawyer, it just seems to me that my life may have been [further] screwed up without even asking me is all. How dare a bunch of bureaucrats and lawyers try to decide what we can and cannot write? Screw them. Abolish the patent system, I say. 8-(

Hartmut Pilch said...

Art 52 .2d EPC can be reworeded in two simple rules: (1) flatly reject program claims and other claim objects that consist of nothing but elements of a programmed universal computer, (2) if there is still no decision, perform some careful, difficult assessments on whether there is a contribution to a field of applied natural science, i.e. a "technical invention".

Courts may find it elegant to decide "against flat rejection" and in favor of "carefully balanced reasoning", but by doing so they create legal insecurity, and in this case, utter confusion about the meaning of the law.

See also http://eupat.ffii.org/07/p2parl/epo
on why this is the straightforward interpretation and other interpretations will tend to make no sense of Art 52 EPC.

With the Macrossan and other decisions, also some German ones, it seemed that courts were getting it.

Hopes that Europe could do significantly better than the US on software patents have been refuted by reality. Even if gradually some of the worst excesses are tempered, the result is still basically the same. Even some of the supposedy "technical" MP3 patents are not significantly better than Amazon 1click in their obscene breadth, and the light-weight economics of the software sector speaks against the kind of heavy-weight regulation and sophisticated doctrines that the UK High Court seems to be again favoring.

TWW said...

I agree with AngryTechie. If this is the way things are going to be run, we as a society would be better off without patents altogether. We can't allow dim-witted old judges to make up laws as they go along. He should be sacked.

Anonymous said...

You know, those who are quickest to find others dim-witted are fastest to reveal something about themselves. David Kitchen dim-witted? David Kitchen old? Does tww know David Kitchen? How old is tww? He certainly knows nothing of patent law. His anti-patent colleague Pilch gives the nod to Macrossan, and Astron has no effect on Macrossan so, tww, why the outrage? If tww doesn't know anything about patent law, he should leave it to those who do, those who represent the patent owner one week, and the infringer the next.

Anonymous said...

In my opinion the only thing this decisions says, is that *if* a claim to a method "for doing process X on a computer" is allowable, *then* a claim to a computer readable medium storing a program "for doing process X on a computer" is also allowable.

The decision says nothing about the conditions that process X must fulfill to be allowable. Maybe process X is fully implemented in software, maybe it is software interacting with novel hardware, etc. These conditions were and are still governed by Aerotel/Macrossan.

If you are of the opinion that a process X that can be implemented fully in software should by definition not be allowable, then that is a valid opinion. However, it is unrelated to the present judgement.

More relevant to the present judgement is whether a computer readable medium that by itself does not do anything and only has the "potential" to achieve an effect, namely when it is inserted into a computer, can still be an allowable invention because of that "potential" effect. You could say that the computer is an essential element of the invention (it is required for achieving the effect), so it should be part of the claim.

As I understand it, the Court now decided that, following the EPO case law, the computer does not have to be part of the claim (and that A/M did not decide anything on this particular point).

A claim to a computer readable medium makes it easier to enforce the patent, as a computer readable medium directly infringes the patent. With just a claim to the method you would have to show that there is an intention to use the medium for infringing on the method claim. I wonder however if that would be such a bad thing? In par. 9 the Court notes that contributory infringement gives no protection against the production and sale of programs in the UK intended for use abroad. But is it not up to the legislator instead of the Patent Office to decide if and how there should be such protection?

Another thing I wonder about: what type of claims are now allowed? Must they be claims to a "computer readable medium carrying a program" or can they be claims to a signal embodying the program as well? According to par. 2 the examiner rejected claims to programs on suitable storage media. I'm not convinced that there should not be a difference between claims to 1. a computer program, 2. a computer readable medium carrying a program, and 3. a signal embodying a progam.

Anonymous said...

Ah, a claim to a "signal". Nice point, that contributor. Not patentable in USA, right? In re Nuitgen, was it? Even though USA allows everything else. Where's the logic in that? So, who's for evolving towards a blend of Steinbrener and A/M, to lead us to a coherent CII position in Europe that is consistent with the EPC. Not just Art 52 but also the Art 69balance between "fair" protection for the technological innovator and legal certainty for the public caught by the other fellow's patent?

Anonymous said...

The problem is not where the boundary is drawn today. The problem is that the boundary creeps.

There is continual assymetic pressure being applied through the legal system via the appeal of successive patent decisions by unsuccessful applicants who have the incentive, money and resources to appeal (and do so frequently). However there is noone in court fighting for the position of angrytechie.

Judgements like this one that partially cave in to that pressure incrementally move the boundary towards allowing patents on software. As a consequence of many such judgements the clear and unambiguous language "software as such is not patentable" somehow comes to mean "all software is patentable if you just use the right words in the application" leaving people like angrytechie who are directly impacted tearing their hair out in frustration.


This judgement is very damaging. The clearest and strongest position for the patent office to defend is that anything which is purely software is simply rejected. This judgement prevents the patent office from making such a clear distinction. It puts us back in the murky world of maybe's where software 'might' be patentable under some vague ill defined circumstances, and all rejections of software must therefore be individually justified. It will be impossible for the patent office to draw a clear line of demarcation in such uncertain ground, and if this judgement holds, future court cases by unsuccessful applicants will rapidly shift the boundary back to the situation where virtually any and all software can be patented willy nilly so long as the right weasel words are included in the application.

Anonymous said...

The problem, you say, is unchallenged "creep". Now a lot of patent owners might agree with you that there is creep, but the sort of creep they have in mind is reverse creep, that of the EPO Boards of Appeal creeping back towards tougher standards for patentability, at the instigation of Opponents. AngryTechie and his ilk should file oppositions, like the Greens do. The EPO Appeal Boards have no obstacles in the way, if they come to the view that they will creep back. It's simply a public policy debate and, with no Binding Precedent restraining the Boards, they can creep any way they choose, helped by 1)opponents like AngryTechie and 2) Industry opponents inconvenienced by the software patent in question.

Anonymous said...

I'm not going to weigh into this argument except to say that the Anon who posted "the only thing this decisions says, is that *if* a claim to a method "for doing process X on a computer" is allowable, *then* a claim to a computer readable medium storing a program "for doing process X on a computer" is also allowable" clearly knows what he/she's talking about. Well done that (wo)man for trying to explain this to the doom/gloom merchants in a simple and logical way.

Wish the UKIPO had a similar level of understanding...

Gerontius

Anonymous said...

Some very good points made above about a consistent European view and what the decision is really about. For AngryTechie and like-minded others though, just please consider some basics for a moment; is software 'technical'? If it's not then why do software designers call themselves 'techies'? I doubt if many people would seriously consider software not to be technical, in which case, just like every other form of technical subject matter, why should software not be patentable? The issue isn't fundamentally about stopping others from copying a programmer's efforts (after all copyright can do that in a more limited way), but about excluding technical subject matter (computer programs in this case) from having the same protection as other inventions. If the software in question is novel and inventive (ie non-obvious), then why should such a technical invention not be deserving of patent protection just like any other inventive technical subject matter? Society decided long ago that inventors could be rewarded for inventions by way of patents, so why should a particular class of inventor be denied such protection? Muddying the waters of this debate by talking about specific cases which arguably should or should not have been granted patent protection doesn't move the debate forward at all.

Anonymous said...

>why should software not be patentable?

Let me try to answer that.

Firstly, because all software is innovative by definition. As programmers we spend our lives solving problems. We do this hundreds of times a day. Each time we do it, we take the risk that our particular solution has been invented by someone else before, and patented. It is unreasonable to expose us, as a profession, to this level of uncontrollable risk. (Have you ever tried to get insurance against it? You can tell that risks are unacceptable when insurers won't offer you cover.)

Secondly, because software is highly abstract, and this makes it almost impossible to decide whether two separate innovations are equivalent. Is an entity the same thing as an object? Who decides, and how?

Thirdly, because patent protection for software inhibits innovation and inhibits competition. The counter-argument that patents encourage innovation relies on the idea that people won't invest in innovation unless they can achieve a return on their investment by being given a monopoly. There is zero evidence that this logic works in the software business. On the contrary, the market is highly distrustful of a technology that's available only from one supplier. The only companies making money from software patents are liquidators trying to make something out of ideas that weren't successful in the market.

Fourthly, because patenting can't distinguish the big ideas, which are so fundamental that human progress would be slowed down if they were restricted, from small ideas which every competetent programmer comes up with every day, from middling ideas - the kind that unusually good programmers have three times in their career - which are the only possible territory where patenting could be justified.

Michael Kay
http://www.saxonica.com/

Anonymous said...

Seems to me that Michael Kay's eloquent contribution makes the EPO's case. Software as such is Statute(EPC)-barred from patentability in Europe. All the innovation in the world helps patentability not one jot, unless the innovation expressed in the patent claim corresponds to the solution of a problem that is (for the EPO) technical. In Europe, software that does no more than solve a problem in marketing or accountancy or derivatives trading can't form the basis of a valid patent. How much of Michael Kay's innovative software delivers an EPO-certified "technical effect"? How much of it takes him outside the EPC-excluded zone of "software"? None, I would think, eh Mr Kay? In that case, it can't infringe any valid patent claim.

Anonymous said...

The link to the judgment is not working - for me at least.

David said...

Sorry about that. It worked for me just now, but perhaps you need to go to the IPKat groups page and find it from there, if the link doesn't work directly.

Anonymous said...

Who ever you are who stated this

[Thirdly, because patent protection for software inhibits innovation and inhibits competition.]

This is wanted by the big companies because they can strangle the smaller startups

[The only companies making money from software patents are liquidators trying to make something out of ideas that weren't successful in the market.]

Not so...everyone in the western world and some in the developing third worlds by the time they die will know of the following companies..

1. IBM
2. Microsoft
5. Google

Because they are untouchable! They won't go because they are the real powerbase...forget your presidents and so forth...these guys provide the money and they control the information! Patents provide a superb control mechanism that keeps them king of the castle. Every line of code that is written by smaller, more innovative, companies can be challenged at any time when they see fit...

Its control of the information that they want. The same way that banks control countries. Through the money...you control IT through the information...you control those 2 and you can do what the hell you want (on this earth only...don't forget that!?)

What we have to worry about is the actual people who want this kind of control...why do they envy it so much? Where are they going with it? That is what we should be worrying about.

Anonymous said...

"Patents provide a superb control mechanism that keeps them king of the castle". In your over-heated imagination maybe, anonymous. How so, exactly? What's this "challenge" you mention. Since when was writing code an act of infringement of a patent claim? Are you saying that patents are the only thing stopping little companies from rising up and confiscating market share from IBM or MS? Dream on, or explain better.

Cristian said...

Perhaps I may have over reacted to David's initial post and have to confess to having made my comment based on David's summary, and not a reading of the case report. It would be interesting if we can actually agree what the case does and does not do to the current status of software patents in the UK.

Anonymous said...

Cristian, this judgment does absolutely nothing to the current status of granting software patents in the UK. Nothing that was not previously patentable will now be patentable and vice versa. The judgment simply overturns some very poor logic on the part of the UKIPO which ran counter to the last 15 years of UK court of appeal case law.

I could yet be proved wrong since the UKIPO don't appear to have commented yet. But this is essentially a judgment about how the UKIPO treated the first step of the four-step test, and the meat of the test remains unchanged so I don't see how the substance of patentability can be changed by this judgment.


The judgment does change the playing field when it comes to software patent infringement. Now a person selling the computer program can be a direct infringer whereas before it was only the end user of that code. Of course, programmers will complain that this works against them, but it's actually good for the average home user of, say, Windows Vista. In the event that Windows is found to infringe a patent, Microsoft can be pursued directly and the 101 million computer users can heave a sigh of relief. Little guy wins, big guy loses - completely the opposite of what most people seem to think.

Having said that, none of the patents that were the subject of the decision are the sort of "mere data processing" patents that are likely to form part of the Windows operating system. They all look like really solid inventions to me (except possibly the chip design one) which no ordinary computer programmer would ever infringe because the invention lies in the useful end result or an underlying advance (eg using less printer toner or modelling human skin), not the programming.


One other comment: I find it interesting that a common complaint about software patents is that clever attorneys can get the patent just by wording it in a clever way. This judgment shows that such a complaint is without merit: the substance matters, not the way in which you claim it. I'll be the first to admit that good writing helps, but it will never get you all the way at either the EPO or the UKIPO.

Gerontius

Cristian said...

The judgment needs careful analysis but on a first reading I would have to say that paragraph 51 is probably the key to understanding the scope of the judgment:

"In all these circumstances I have reached the conclusion that claims to computer programs are not necessarily excluded by Article 52. In a case where claims to a method performed by running a suitably programmed computer or to a computer programmed to carry out the method are allowable, then, in principle, a claim to the program itself should also be allowable. I say "in principle" because the claim must be drawn to reflect the features of the invention which would ensure the patentability of the method which the program is intended to carry out when it is run."

Kitchin's reasoning is based on the conclusion that Macrossan did not completely exclude computer program claims but left the door open in certain limited cases. Ultimately this is a question of interpretation, whether it be the interpretation of Article 52 or past decisions at both a UK and EPO level. Judges faced with the same facts and looking at the same past case law have arrived at often very different conclusions in the last few years. I cannot say the interpretation on this point and others in the judgment is wrong, because the ambiguity in the EPC creates the possibility for varying interpretations. However, any decision which results in the possibility of program claims, even in limited cases, may have a significant economic impact, as noted in Macrossan, and at this stage it would be preferable if clarity was provided by a democratically elected legislature and not through the continued reliance by the UK judiciary on a long line of EPO decisions decided by insitutions that have no direct accountability to the public.

Anonymous said...

On my view, the core issue lies in Jacob not clearly distinguishing between the EPO practice before and after the two IBM decisions (and, thus, between the approach followed at the UKIPO before and after those decisions): the technical effect approach “is the approach (with the rider) adopted by this Court in Merrill Lynch. It has been followed in the subsequent decisions of this Court, Gale and Fujitsu. The approach (without the rider as an express caution) was that first adopted by the EPO Boards of Appeal, see Vicom, IBM/Text processing and IBM/Data processor network”. Jacob retains to be bound by his precedents in Merrill Lynch, Gale and Fujitsu.
Although it is clear that the four-step test catapulted the UK back to the way of thinking of Merrill Lynch (Gale and Fujitsu) and Vicom, a clear distinction should have been made between the EPO doctrine before and after the two IBM decisions. It is, in fact, after the IBM decisions at the EPO that also the UKIPO started to admit computer program products claims, while they were not allowed before, under the Merrill Lynch “technical contribution” approach. On these grounds, it appears to me that, under the four-step test, computer program products should not be admitted.

Anonymous said...

What the above anon states as "fact" is actually (half) wrong.

"It is, in fact, after the IBM decisions at the EPO that also the UKIPO started to admit computer program products claims, while they were not allowed before, under the Merrill Lynch “technical contribution” approach."

Merrill Lynch said nothing about computer program claims, so it's not a question of "not being allowed to" it was a case of playing it safe and deciding not to allow them until IBM pointed out the illogicality of accepting claims to a computer programmed to achieve a technical result and not the computer program itself - extolling form over substance in a way repeatedly denigrated by the UK courts.

On my reading of the recent judgment, the judge was influenced more by the UK court judgments than the EPO decisions: the EPO may have first suggested explicitly that computer program claims were a possibility, but the Court of Appeal judgments certainly support that viewpoint.

Relevant quote follows

"31. In my judgment the United Kingdom Patent Office was right to revise its practice as it did. It seems to me to be the logical consequence of the dual approach that claims to computer related inventions must be considered as a matter of substance not form, as Genentech, Merrill Lynch, Gale and Fujitsu so clearly establish, and that what is needed to make an excluded thing patentable is a relevant technical contribution. If a program makes a conventional computer operate in a new way so as to deliver a relevant technical contribution then it seems to me to be wholly artificial to say that the effect is delivered by the computer but not the program. If, as these cases decide, a conventional computer programmed with such a new program is patentable because it is no longer a computer program as such then, in my judgment, the same reasoning must apply to the program itself."

Gerontius

Mazhar Hussain Shah said...

The judgment does change the playing field when it comes to software patent infringement. Now a person selling the computer program can be a direct infringer whereas before it was only the end user of that code. Of course, programmers will complain that this works against them, but it's actually good for the average home user of, say, Windows Vista. In the event that Windows is found to infringe a patent, Microsoft can be pursued directly and the 101 million computer users can heave a sigh of relief. Little guy wins, big guy loses - completely the opposite of what most people seem to think.

Having said that, none of the patents that were the subject of the decision are the sort of "mere data processing" patents that are likely to form part of the Windows operating system. They all look like really solid inventions to me (except possibly the chip design one) which no ordinary computer programmer would ever infringe because the invention lies in the useful end result or an underlying advance (eg using less printer toner or modelling human skin), not the programming.


One other comment: I find it interesting that a common complaint about software patents is that clever attorneys can get the patent just by wording it in a clever way. This judgment shows that such a complaint is without merit: the substance matters, not the way in which you claim it. I'll be the first to admit that good writing helps, but it will never get you all the way at either the EPO or the UKIPO.

Gerontius

Subscribe to the IPKat's posts by email here

Just pop your email address into the box and click 'Subscribe':