A total of 18 non-confidential submissions were received in response to IP Australia’s consultation papers. Once again, a number of these submissions argued against changes in the law, notwithstanding that the Government has already decided to implement the various Productivity Commission recommendations, and doing nothing is therefore not an option that is open to IP Australia. Fifteen non-confidential submissions provided comments on the options for amending the inventive step requirements, of which 14 did not support making any changes. Fourteen non-confidential submissions were received in relation to introducing an objects clause, of which just six submissions were in favour, and eight did not support the change.
There was less stakeholder interest in Crown use and compulsory licensing, however among those that did make submissions on these issues there was somewhat greater support for change. Of seven non-confidential submissions relating to Crown use, five supported amending the law. In relation to compulsory licensing provisions, three out of eight non-confidential submissions supported change.
Crown use and compulsory licensing provisions have rarely been invoked in Australia (and no compulsory licence has ever been granted). Furthermore, the proposed amendments are unlikely to result in any increase in the use of these provisions. The relative lack of interest in these areas is therefore not very surprising.
However, the proposed changes to the inventive step requirements, and the preferred wording of the objects clause, could have the more profound effect of introducing a ‘technology’ requirement into Australian patent law, creating tension with previous statements of the Full Bench of the Federal Court of Australia rejecting the proposition that patentable inventions be required to fall within an ‘area of science and technology’.
While the majority of Productivity Commission recommendations addressed by the draft legislation relate to PBRs and trade marks, the most significant patent-related provisions are those implementing the abolition of the innovation patent system. The proposed approach would see a fairly gentle transition, with applicants given a year’s notice within which innovation patents would still be granted under the current system, followed by eight years during which all existing innovation patents would gradually cease or expire, and restrictions would be imposed on new applications and the certification of innovation patent claims.
When the draft legislation was published, I argued that the proposed transition could be viewed as too gentle, considering the Productivity Commission’s relatively strong objections to the strategic use of divisional innovation patents, and the uncertainty caused by uncertified claims. This was, however, very much a contingent view – i.e. given that the innovation patent system is to be abolished based upon the Productivity Commission’s recommendation, the abolition should be implemented in a way that most-effectively addresses the Commission’s reasons for making that recommendation.
In fact, I would prefer to see the innovation patent system given a second chance, in an improved form. However, since the Government has accepted the Productivity Commission’s recommendation, and the consultation was directed to its implementation, rather than being a further opportunity to canvas objections to this decision, I did not make a submission. This did not, however, prevent a number of others from taking the opportunity to once again express their opposition to abolishing the innovation patent system.
Unsurprisingly, these repeated objections fell on deaf ears, with IP Australia focussing in its response on the few submissions that actually addressed the merits of the proposed legislation. On this basis, it appears at this stage that – absent a last-minute reprieve at the parliamentary level – the process of abolishing the innovation patent system is likely to proceed in substantially the manner set out in the draft legislation.
Within the realm of conventional ‘small molecule’ drugs (i.e. setting aside, for present purposes, recent developments in biotechnology and biologics) four broad categories of patent protection can be identified.
New compounds. When an innovator discovers or synthesises a compound that was not previously known to exist and/or to have any useful therapeutic effect, a patent may be obtained for the compound itself.
Formulations. While it is one thing to identify a biologically-active compound, it is often another altogether to find a safe and effective way of delivering it to a patient. When an innovator develops a new and/or more effective formulation (e.g. a particular tablet or other oral dosage form), a patent may be obtained for that formulation, even if the active ingredient is well-known and no longer patentable.
New therapeutic effects. When an innovator discovers that a compound already known to have one or more therapeutic effects has a further, and unforeseen, therapeutic use (e.g. in the treatment of a different disease or condition), a patent may be obtained for a method of using or making the compound specifically for the new therapeutic purpose.
Improved methods of treatment. When an innovator discovers that a known compound with a known therapeutic use can be made even more effective, e.g. by combining it with other compounds or therapies, or using a different dosing regime, a patent may be obtained for the new method of treatment.
Of course, a patent will only be validly granted in any of these cases if all of the legal requirements for patentability – novelty, inventive step, sufficiency of description, and so forth – are all satisfied. These requirements are intended to balance the rewards available to innovators against the rights of the broader community to access knowledge and to engage in free competition.
From a policy perspective, getting the balance right is particularly important in the case of pharmaceutical products. If it is too difficult to obtain a valid patent, there may be insufficient incentive for companies to invest billions of dollars in new drug development. On the other hand, it is important to keep in mind that, one way or another, it is the wider community – either individually, or through taxes in countries where healthcare is substantially subsidised by government – that ultimately pays for that development, through the higher prices charged for patented drugs. Allowing patents to be granted too easily therefore may therefore represent a significant social cost.
A major component of the cost of bringing a drug to market is the need to conduct extensive clinical trials in order to prove the effectiveness and safety of the drug, and thus to obtain regulatory and marketing approval. Such trials typically take years to complete, following the initial discovery of a new compound, or of a new use for an existing compound. However, a patent application must be filed as early as possible to ensure that any available protection is secured in the event that the trials are successful.
A recent appeal decision of a Full Bench of the Federal Court of Australia sheds light on how the balance between an innovator’s need to file early, and the community’s right to receive a full disclosure of the invention, is struck in the case of a patent for a new therapeutic use of a known compound: Warner-Lambert Company LLC v Apotex Pty Limited (No 2)  FCAFC 26. The court has confirmed that the fact that the clinical trials required to establish an effective and safe dosage of a drug for a new therapeutic use may be ‘complex, time-consuming and expensive’ will not render invalid a patent that was filed prior to conducting this essential research.
The new merger is significant in terms of its impact on clients’ choice of patent attorneys in Queensland. According to the Register maintained by the Trans-Tasman IP Attorneys Board (TTIPAB), FAKC and Cullens collectively employ 26 patent attorneys (and one trade marks attorney) registered at Queensland addresses, out of a total of 93 patent attorneys registered in Queensland. Following the merger, the largest patent attorney practice in Queensland will be Davies Collison Cave (five attorneys registered in the state), followed by Eagar & Martin, IP Gateway and Pizzeys (all with four Queensland patent attorneys, although Pizzeys also has six patent attorneys based in Canberra). Of these, Pizzeys is itself a member of the IPH group (though not – yet – merging with S&F) while Davies Collison Cave is Australia’s largest IP firm and a member of the QANTM IP Limited (ASX:QIP) group. Thus any client in Queensland wanting to work with a mid-tier patent attorney firm will most likely need to look interstate.
I predict that this will not be the last merger of firms within the three publicly-listed ownership groups (the third being Xenith IP Group Limited, ASX:XIP). While it is commonly believed that the scope for mergers may be limited by the need to avoid conflicts of interest between the clients of the separate firms, my own conversations with people within the groups suggest that this is not necessarily viewed as a major obstacle.
Furthermore, the updated Code of Conduct for Trans-Tasman Patent and Trade Marks Attorneys 2018, which came into effect on Friday 23 February 2018, creates a presumption of non-independence of firms within an ownership group. The Code also imposes new obligations on the firms within a group to inform clients of the arrangements, and to obtain informed consent, in writing, from clients on both sides of any contentious matter (e.g. a patent opposition) represented by different firms within the same group. It is my view that, in some cases at least, these more-onerous obligations may tip the balance in favour of merging firms, rather than maintaining separate operations.
The real question, then, is not if there will be further mergers, but who will be next?
I recently wrote about IP Australia’s new Engaging an Attorney Toolkit, which is an online ‘guide on how, why, what and when to engage your patent attorney’ that is intended primarily to assist people and businesses with minimal knowledge and experience of the patent system in preparing to engage with an attorney. The toolkit includes a section entitled ‘6 Myths about getting a patent’, the first of which is that ‘a patent will cost me hundreds of thousands of dollars’. The myth-busting reality, according to the toolkit, is that ‘prices vary, but the cost of drafting and filing an initial patent application (known as a provisional application) for your invention typically costs somewhere between $3000 and $6000.’
Is this information correct? Or – as one commenter suggested – is a price range of $3000-$6000 liable to create unrealistic expectations, particularly for those prospective clients looking to engage with a major attorney firm?
The toolkit’s estimate is not, in fact, unreasonable, in the sense that the median cost to have a patent specification drafted by an Australian patent attorney is almost certainly somewhere in the range given. But it is also true that costs for a substantial proportion of all specifications drafted in Australia would fall outside this range. There are individual attorneys and small firms that might be willing to draft a patent specification for a technically simple invention for as little as $2000. At the other extreme, an experienced attorney at a top-tier firm might end up charging $12,000 or more for a specification for a complex invention in a sophisticated and challenging field of technology. Since many attorneys charge by time, and hourly rates vary significantly with experience and across different firms, there are no hard-and-fast rules about what a patent application may cost.
There are, however, hard-and-fast rules about what a patent attorney must tell a client up-front about who they are, how they work, and how much they will charge. It is timely to mention this, because the new Code of Conduct for Trans-Tasman Patent and Trade Marks Attorneys 2018 just came into effect, on Friday 23 February 2018. All registered patent attorneys in Australia and New Zealand are subject to the Code, which requires (among other things) that they inform each new client, in writing:
that they are, in fact, registered as a patent attorney, and are bound by the Code;
that they have appropriate competency to perform the work required by the client, including suitable expertise in the technology of the invention;
whether the attorney works for an incorporated firm and, if so, whether it is privately or publicly owned;
the procedures, timing and estimated cost of doing the work required by the client.
(With regard to registration, prospective clients should also be aware that Australian law effectively prohibits anybody who is not a registered patent attorney from drafting patent specifications on behalf of others. In particular, under subsection 201(1) of the Australian Patents Act 1990 only a registered patent attorney or legal practitioner may act as a patent attorney in Australia, and section 202 further provides that, other than in very specific circumstances, ‘a legal practitioner must not prepare a specification, or a document relating to an amendment of a specification’.)
So any prospective client can be assured of being made aware of anticipated costs before any work commences, and can, if they wish, shop around for the best deal. But how is a newcomer to the patent system to determine what is reasonable, given the very wide range of cost estimates that may be provided?
Sorry – it seems I have that wrong. According to IV’s profile of co-founder Nathan Myhvold, the company is actually ‘building a market for invention’ to enable inventors to ‘realize the value of their ideas’. It ‘manages one of the largest and fastest growing intellectual property (IP) portfolios in the world, with more than 40,000 assets and more than $6 billion in total committed capital’ and its investors include ‘many of the world’s most innovative companies and renowned academic and research institutions.’ All of which actually sounds kind of like a good thing, and not at all troll-y. It is also notable that Detkin’s original definition would encompass most universities and research institutes.
So, where does the truth lie? Somewhere between the two extremes, no doubt, as technology analyst Roger Kay has eloquently explained in an article on Medium this week, Where did the Patent Troll Narrative Come From? At one end of the spectrum, there are certainly genuinely predatory ‘trolls’ using the patent system to extort payments from companies that lack the resources to fight back. At the other end, however, are big tech companies, like Intel, Google, Facebook, and Apple, which have adopted a strategy of ‘efficient infringement’ – the practice of using a technology that infringes on someone’s patent, ignoring the patent holder entirely, and when (or if) the patentee decides to sue, tying them up in court by challenging the patent’s validity. As I have argued in the past, slapping the positive word ‘efficient’ on the front does not alter the reality that this is also an abuse of the system, and of the infringers’ superior economic power.
In other words, it’s complicated. As Kay notes:
Among combatants in the patent wars, entities are divided along multidimensional lines: practicing vs. non-practicing, dominant vs. upstart, technological vs. financial, inventing vs. acquiring, licensing vs. using internally, R&D-oriented vs. manufacturing-oriented, invention-focused vs. product-focused. While some of these creatures are truly odious, using the word “patent troll” to describe any of them would be to allow Intel and its allies ownership of the narrative. There are no hard and fast rules to separate them into good and bad buckets.
The latest Patent Litigation Year in Review report, from legal analytics company Lex Machina, provides some insights pertinent to the troll narrative, and in particular the impact of changes in US patent law driven, in large part, by lobbying from the ‘efficient infringer’ constituency which, as noted in this IAM Media article, ‘invested large sums in spreading a narrative around the problems posed by “patent trolls” that has been used to justify the need for a re-engineering of the US system to make it less friendly to all rights owners.’
Lex Machina’s analysis shows that since the commencement of the US patent law reforms introduced by the America Invents Act (AIA), rates of patent litigation have been in steady decline in real terms. Furthermore, while the list of top plaintiffs remains dominated by non-practising entities (NPEs), in 2017 two pharmaceutical companies entered the top ten, with two more filling out the top 15. And while headlines tend to be captured by a small number of very high awards of damages against big infringers, the reality for most plaintiffs is sobering. Just 11% of all cases terminated since 2000 reached a final judgment, with around three-quarters settling. While patentees are victorious slightly more often than defendants (around 60/40), compensatory damages are awarded in less than half of the cases won by plaintiffs, and for those cases in which ‘reasonable royalty’ damages were awarded during the three years up until the end of 2017, the median amount was just US$4.4 million – perhaps barely enough to justify litigation in a jurisdiction where the usual rule is that each party must bear its own costs of the proceedings.
While this might be bad news for the genuine trolls in the system, it is at least as bad – if not worse – for individual inventors, research groups, universities, and other innovators who are better-placed to invent than they are to commercialise their inventions, and therefore rely on licensing to secure a return from their efforts.
IP Australia – the government authority responsible for (among other things) examining and granting Australian patents – has just published its Engaging an Attorney Toolkit, an online ‘guide on how, why, what and when to engage your patent attorney.’ The toolkit is the result of a project undertaken in the second half of 2017, with the assistance of external branding and communications consultants, having the aim of dispelling myths and closing knowledge gaps around patent protection. It is intended primarily to assist people and businesses with minimal knowledge and experience of the patent system in preparing to engage with an attorney.
In the course of the project, online surveys and interviews were conducted with various stakeholders, including patent attorneys (full disclosure: I was one of the patent attorneys interviewed), to identify what new prospective patent applicants know, do not know, and should know, about the process of obtaining a patent. Naturally, IP Australia is primarily concerned with assisting the public, and is hoping that the toolkit will help to reduce costs and make it easier and more attractive for innovators with limited understanding of the patent system to engage with an attorney. If it is successful, however, the toolkit will also benefit attorneys by creating better-prepared and more knowledgeable clients, reducing the time and effort often required to educate them about what to expect from their attorney, and from the patent system.
…for the overwhelming majority of prospective applicants, ‘seeking assistance from a patent attorney’ is not something that they ‘may consider’. It is an essential step without which they might as well flush their application fees, and whatever their own time and effort is worth, straight down the toilet!
Compared to this short-term fix, however, the Engaging an Attorney Toolkit is a huge leap forward. Not only does it contain valuable information, in an accessible format, but the very fact that it now exists sends a clear message to prospective applicants about the importance of seeking professional assistance. For the first time, ‘Engaging an attorney’ is now a headline topic on the main page on applying for a patent, along with a link to the toolkit. With just one small caveat (of which, more later) I would have no hesitation, if I were still working as a private-practice attorney, in referring prospective clients to the toolkit in the expectation that it would make the early stages of our engagement run more smoothly.
In the first part of this series of articles, I argued that invention is inherently a creative act, and that since machine learning systems – however impressive or surprising their achievements – are incapable of human-like intelligence, reasoning, agency, or creativity, they therefore cannot ‘invent’. I acknowledged, however, that machines are certainly increasingly involved as ‘assistants’ in the process of invention. In the second part of the series I argued that where the result is a patentable invention there must always be at least one human inventor. In this final part, I want to look at how we should go about identifying the inventor(s) in any given case of machine-assisted invention.
There are, in fact, three main aspects of machine learning technology in relation to which inventions may arise. The most rarefied of these is in the underlying machine learning algorithms and architectures themselves. Comparatively speaking, very few people work in this area, and for the most part they are to be found in universities and research centres.
Secondly, inventions may arise through the application of machine learning technology to solving problems and/or producing new results, products, and services. I expect that this is currently the most common type of machine-learning-related invention, particularly given the wide range of software tools now available to assist programmers in implementing the underlying algorithms.
Thirdly, there are ‘machine-assisted’ inventions, which are generated wholly or in-part by machine learning systems. Currently, such inventions are relatively rare, considering that few applications of machine learning are actually directed to the generation of new technologies. However, as machine learning increasingly finds its way into computer-aided design and engineering applications, this may change.
In all cases, however, I would argue that it is possible – and, indeed, necessary – to identify one or more human inventors (and no machine inventors). To suggest otherwise is, in my view, to misunderstand the true nature, and limitations, of machine learning systems. In a letter to shareholders, published in 12 April 2017, Amazon CEO Jeff Bezos wrote what is possibly the most succinct and jargon-free summary of what distinguishes machine learning systems from ‘traditional’ programming:
Over the past decades computers have broadly automated tasks that programmers could describe with clear rules and algorithms. Modern machine learning techniques now allow us to do the same for tasks where describing the precise rules is much harder.
The evolution from ‘traditional’ programming to machine learning is not as dramatic as some of the hype might lead us to believe. Instead of coding the ‘rules’, machine learning developers now build systems that are able to capture and generalise from patterns that exist in their input data. These systems still operate using rules and algorithms – but now these algorithms determine how they go about doing the capturing and the generalisation, rather than how they produce the final result. Viewed at an appropriate level of abstraction, then, little has changed, except for the power and scale of our machines.
In view of this, I argue that when it comes to machine-assisted inventions, inventorship will generally arise from successfully applying a machine learning system on the path to achieving an inventive result, even in cases where the software and/or hardware employed may have been developed or supplied by someone else.
On 17 January 2018, a panel of the European Patent Office (EPO) opposition division wholly revoked a patent co-owned by the Broad Institute (‘Broad’) relating to CRISPR/Cas9 ‘gene editing’ technology. The European patent in question, number EP2771468, is entitled ‘Engineering of systems, methods and optimized guide compositions for sequence manipulation’, and is a European equivalent to US patent no. 8,906,616, which is one of the key Broad patents involved in the US patent interference dispute with the University of California (UC) – which is currently on appeal to the US Court of Appeals for the Federal Circuit. As is usual with European opposition proceedings, the ruling was issued immediately at the hearing (which had originally been scheduled to continue for two further days). A full written decision providing detailed reasons is likely to be a few weeks away.
Technically, the basis for revocation in the final decision is likely to be lack of novelty and/or inventive step of all claims of the Broad patent. However, the underlying reason for Broad’s failure to defend its patent is a loss of priority. In particular, the EPO panel determined that Broad was not entitled to claim priority from four of its earlier US provisional applications, including the earliest filing, US provisional application no. 61/736,527, which was filed on 12 December 2012. This loss of priority was fatal to the patent, as a result of a number of publications – including Broad’s own – that occurred subsequently, but prior to the full application’s filing date of 12 December 2013.
While this decision is obviously pivotal in the ongoing disputes between Broad and UC over ownership of foundational patent rights relating to CRISPR/Cas9 technology, it also provides an object lesson and a timely reminder of essential requirements for valid priority claims for patent applicants around the world, in all fields of technology. While the circumstances of Broad’s case are somewhat more complex than most patent filings, what has happened to it here is not at all specific to the particular invention at issue. In a nutshell, the problem was that, at the time of filing international application no. PCT/US2013/074819 (from which the European patent is derived), the named applicants – Broad, MIT and Harvard – did not collectively own all of the rights necessary to claim priority from the earlier provisional applications. More particularly, the provisional applications named a researcher from Rockefeller University as a co-inventor/applicant, yet neither the researcher nor Rockefeller was named as an applicant on the international application. Nor had the named applicants received any assignment from Rockefeller of the right to claim priority in the international application.
In fairness to Broad, the US national law relating to ‘internal’ priority (i.e. claiming the benefit of a US provisional application in a subsequent non-provisional application) is less strict, and focusses on the substance of the invention actually claimed in the later application rather than on a distinct right of priority. But this can hardly be an excuse for ignorance of the international position, especially when the stakes are so high. Broad has issued a statement, arguing that the EPO decision is based on a technicality, and asserting that it is ‘inconsistent with treaties designed to harmonize the international patent process, including that of the United States and Europe’. I disagree. Not only is this not the first time that the EPO has applied these rules in relation to priority claims, but it is not the only adjudicating body to have determined that they are, in fact, required by the international treaties in question.
Broad has vowed to appeal. In the meantime – and in anticipation of the likelihood that any appeal will fail – all international applicants would be well-advised to ensure that they are clear on the ownership of the right to priority at the time of filing. At the end of this article, I set out some guidelines for avoiding the troubles that Broad has encountered, not only at the EPO but at any international patent office.
In addressing this question, it is important to understand that ML systems do not autonomously or independently generate novel outputs. In my view, this is a fundamental error of understanding in Professor Ryan Abbott’s paper ‘I Think, Therefore I Invent: Creative Computers and the Future of Patent Law’, Boston College Law Review, Vol. 57, No. 4, 2016 (also available at SSRN), which I discussed in the first article of this series. Abbott contends, in particular, that ‘machines have been autonomously generating patentable results for at least twenty years and that the pace of such invention is likely increasing.’
This, I have argued, is simply wrong. The difference between a computer that is programmed to play the board game Go, and one that is programmed to learn to play Go is, of course, significant. The former can only make moves that are determined in accordance with its explicit programming, whereas the latter may appear to ‘invent’ new strategies in response to patterns occurring in its training data that have not previously been recognised by human players. But the appearance of invention is not the same thing as actual invention. The ML player is still doing nothing more than following the instructions devised by its programmers. The Google DeepMind AlphaGo system has become the world’s best Go player as a result of years of development, trial, experiment, and experience on the part of its designers. AlphaGo plays as well as it does simply, and only, because that is what it was designed to do. In this sense it is no more ‘autonomous’ than any other computer program.
In this second article I will explain why I believe that in the case of all existing (and currently foreseeable) ML systems which may generate inventions as output, there is always a human inventor. This is consistent with the history and current state of patent law, as well as with the practical and technical reality of ML systems.
Read Full Article
Read for later
Articles marked as Favorite are saved for later viewing.
Scroll to Top
Separate tags by commas
To access this feature, please upgrade your account.