Software Patentability: Is the USA Following Europe?

June 2, 2013

The patentability of software and business methods remains a controversial, and often confusing, issue. Until recently there was a common perception that while ‘anything under the sun which is made by man’ would be patentable in the US, Europe was much more restrictive as to what could be patented. While there is some truth in this, it has always been the case that much software is patentable in Europe, and recent cases in the US courts are in fact painting a picture of both confusion and an increasing keenness to restrict the availability of patent protection for software implemented inventions. 

Applicants for patent protection will in future need to consider more carefully whether their inventions can be protected in the USA, while appreciating that the position in Europe is one of relative certainty, allowing many software-implemented inventions to be protected. 

USA – Birthplace of Business Method Patents 


For over a decade the USA was, in patent terms, the brave new world where inventors of novel business processes could obtain, often very valuable, patent protection. It was common for patents to be granted for novel methods of commerce and novel methods for increasing one’s wealth by employing a particular investment strategy. No matter how abstract the method was, the US Patent Office (USPTO) was generally happy to provide patent protection for methods which were new and not obvious. 

Recent years, however, have seen increasing restriction on what is patentable – or what constitutes ‘statutory subject matter’, to use the US terminology. This shift gained its highest profile as the case of Bilski made its way through the US Appeals Court and on to the Supreme Court. Mr Bilski had filed a patent application for a method of hedging losses in one segment of the energy industry by making investments in other segments of the energy industry, which was refused. In subsequent appeals this refusal was upheld, by the Court of Appeal of the Federal Circuit, and subsequently in the Supreme Court.   

In the Supreme Court’s judgment, the machine-or-transformation test for patent eligibility, in which a process must be implemented with a particular machine or else transform an article from one thing or state to another, was explicitly rejected as the sole test for determining patent eligibility, making clear that in some circumstances the machine-or-transformation test may not be the only indicator of patentability. The Court also rejected exclusion of all business method patents from eligibility. However, it did not provide any clear guidance on how to determine whether a business method relates to patent eligible subject matter or what tests other than the machine-or-transformation test might be appropriate for assessing those cases where the machine-or-transformation test is not appropriate.  

Most recent developments 

While some might have hoped that the period after Bilski would be one of certainty and stability, this was not to be. This was perhaps inevitable given the Supreme Court’s refusal to provide clear guidelines as to patentability. Such a situation comes as no surprise to practitioners in this area in Europe, as we are long accustomed to the latest judicial pronouncement in this troubled area raising more questions than it answers, thereby requiring further judicial guidance in future cases, and giving rise to further uncertainty in the process. And so it is not that surprising that we now have another high-profile case on this issue: CLS Bank v Alice. 

In May 2007 CLS Bank sought to have certain patents relating to a computerised trading platform used for conducting financial transactions in a way that eliminates ‘settlement’ risk, owned by Alice, declared invalid and unenforceable. The patents in question were essentially directed to a form of computer-implemented escrow. Alice counterclaimed for infringement of the patents by CLS. In March 2009 CLS moved for summary judgment, arguing that all asserted claims were drawn to patent ineligible subject matter, and therefore invalid. The District Court denied the motion, essentially because of the Bilski case then pending before the Supreme Court.  

After the Supreme Court’s Bilski decision was handed down, CLS renewed its motion for summary judgment on the question of patentability. The District Court held all of the asserted claims unpatentable. This was in some ways surprising because many of the asserted claims were system claims – and the parties agreed that the method claims were limited to use of a computer – and might therefore have been expected to pass the ‘machine’ part of the machine or transformation test.

Unsurprisingly, CLS appealed to the Court of Appeals for the Federal Circuit – which hears all patent appeals – and that court decided that an en banc sitting was required to again try to bring clarity to patentability. Though laudable in its intention, this move would later prove to be a failure.

Ten judges sat en banc to reconsider the patentability of the claims asserted by Alice against CLS Bank, to try to provide the much longed for, but now more elusive than ever, clarity relating to patent eligibility. Instead of achieving the desired clarity, the judges reached ‘judicial deadlock’, with Judge Newman noting that:

the only assurance [provided by the different opinions of the ten judges] is that any successful innovation is likely to be challenged in opportunistic litigation, whose result will depend on the random selection of the panel

The ten judges’ views as to the patentability of the asserted claims spanned the full spectrum of possible combinations, ranging from all system, method and computer readable media claims being unpatentable to all claims being patentable, with a view in between that the system claims were patentable while the method and media claims were not. Further differences in reasoning for the views resulted in a total of six variations of opinion amongst the ten judges.

As noted by Chief Judge Rader, no portion of any opinion other than the per curiam judgment – supported by only five of the judges – garnered a majority. As noted by Judge Rader:

The court is evenly split on the patent eligibility of the system claims. Although a majority of the judges on the court agree that the method claims do not recite patent eligible subject matter, no majority of those judges agree as to the legal rationale for that conclusion. Accordingly, though much is published today discussing the proper approach to the patent eligibility enquiry, nothing said today beyond our judgment has the weight of precedent.

What is clear, however, is that with half of the judges rejecting the system claims, there has been a swing in the US towards a harsher treatment of patents directed towards business methods. Indeed, the opinion with which the greatest number of judges concurred indicates that, in order for a process to be patent eligible, an ‘inventive concept’ is required in addition to any abstract idea on which the invention is based. In an attempt to explain what is meant by an inventive concept, the opinion indicated that the human contribution must represent more than a trivial appendix to the underlying abstract idea and that: 

Limitations that represent a human contribution but are merely tangential, routine, well-understood, or conventional, or in practice fail to narrow the claim relative to the fundamental principle therein, cannot confer patent eligibility. 

Whilst a requirement for more than an abstract idea is in line with previous case law, that previous case law was clear that the abstract idea restriction was to be interpreted narrowly. Based on this narrow interpretation, it was generally possible to avoid the abstract idea restriction simply by specifying a computer in the claims, meaning that patent eligibility could be achieved in a relatively straightforward way. It seems however that now this may not be enough, or at least, this may not be enough if faced with one of the five judges that sat en banc in CLS Bank v Alice Corp, but may well be enough if faced with one of the other five judges. 

There is, therefore, little certainty around patent eligibility in the US, and indeed, despite efforts of recent years to clarify the position, if anything the American approach is now less clear than ever.

The USPTO has indicated that it will issue new guidance for patent examiners in the light of CLS Bank v Alice. It will be interesting to see what guidance the USPTO can draw from such a divergent and contradictory set of opinions.  

Meanwhile, in Europe… 

In Europe, the position on this issue has not been entirely consistent either. Indeed, while the European Patent Office (EPO) has followed one course, the UK Court of Appeal (binding lower tribunals including the UK Patent Office) has followed its own course.  

The EPO 

Readers may be aware that patent laws in Europe specify a list of ‘things’ that are excluded from patentability (ie are not patent eligible). Included in this list are computer programs and business methods ‘as such’. It is therefore no surprise that European tribunals have spent many decisions trying to give some meaning to the ‘as such’ limitation so as to explain what is and is not patentable. This has caused damaging uncertainty for European applicants who often wrongly assume their software-implemented inventions are not patentable when, in fact, many software-implemented inventions can qualify for patent protection in Europe. 

From the earliest of its decisions, the EPO has taken the line that European patents are available for technical – not business – inventions. It initially explained this by saying that a ‘technical contribution’ must be provided if an invention is to be patentable. However, this position developed somewhat in the first years of the 21st century to provide a two stage approach: 

1.      To avoid the statutory exclusion the invention must not be entirely abstract. It must include at least one physical component (eg a computer). As such where a physical component is claimed – whatever that may be – the statutory exclusion is avoided. It follows that a business method invention can easily pass this step by being claimed in a computer implementation. 

2.      However, in order to involve an inventive step – a requirement for patentability – the invention must also provide a technical solution to a technical problem. The mere provision of a business or administrative solution to an administrative problem will not suffice. 

In applying this approach, the EPO has made clear that, in inventions involving business methods, the novel non-technical features (ie the novel business method) can be placed in the definition of the problem to be solved, meaning that in many cases the only solution is to use a computer to implement the business method – something which will in many cases be so obvious that protection for the idea cannot be obtained anyway. 

This approach – while not without its critics – has provided a period of certainty and stability for applicants for patents for computer-implemented inventions at the EPO. While there have been attempts to have the approach reviewed – both by way of legislative change and by way of referral to the EPO’s Enlarged Board of Appeal for review of the law – these have failed, and the EPO seems set to continue in its practice. 

The UK courts 

Concurrently, the UK courts have refused to follow the EPO’s lead, encouraged by the UK Patent Office both when acting as amicus in various appeals and when making submissions in the attempted Enlarged Board referral at the EPO. The UK position resists the two-step approach and insists that in order to meet the patentability requirements an invention must provide a technical contribution. It uses a four-step test to assist in this enquiry, which includes a consideration of whether the contribution (a concept not defined with any great precision) is technical. This is clearly somewhat circular and so it is unsurprising that the four-step test can often lead to the same difficulties. 

The UK courts’ decisions have long made it clear that comparisons between individual cases are often not helpful. While this is undoubtedly often true, it is hardly helpful when one is formulating arguments for or against patentability. Seemingly accepting this, Lewison J (as he then was) decided that some ‘sign-posts’ as to whether an invention involved a technical contribution would be helpful. However, these sign-posts have unfortunately provided little assistance, due to their involvement of comparisons between cases which are rarely entirely legitimate, as noted in earlier cases. This has recently even been observed by Lewison LJ (as he now is) in a recent case where he needed to amend one of his ‘sign-posts’ to show why an invention before him was in fact patentable. 

Although at least currently settled, the UK position involves an approach which many consider to be inferior to that applied by the EPO – an opinion reinforced by recent changes to the sign-posts. It is hoped that the UK courts will soon follow the EPO’s approach, at least to provide greater clarity for applicants – many of whom remain confused as to the position in Europe.  

The Future 

The US courts are clearly showing an inclination to restrict patentability. The judgment in CLS leaves little room for doubt that earlier guidance from the US Supreme Court is being applied to restrict the extent of patentable subject matter. The manner of that restriction has loud echoes of the approach that the European position has settled on – does the invention relate to non-excluded subject matter? And much of the way in which this is applied in CLS leads us to think that perhaps the US Court is nearing the EPO’s approach. 

Further developments surely await us in the coming months and years. While we are perhaps closer than ever to US-European consensus on this difficult issue, the approach that seems most certain to endure is the EPO’s. It might have been controversial and has even been dubbed ‘not intellectually honest’ by a prominent Lord Justice of Appeal, but it seems to be standing the test of time. 

Mark Kenrick is a partner, and David Robinson is an associate, at the Manchester Office of Marks & Clerk LLP. Both specialise in patent matters relating to computer-implemented inventions.