Search the archive.

The potential consequences could be very different in other fields of technology. For the global tech giants, in particular, it may be a short step from computerised invention to machines that also generate the documentation necessary to file for patent protection. Automated patent generators capable of monopolising large areas of technology could become a reality within the decade. If the primary legal barrier to this development is a requirement that an inventor be human, becoming the first, and possibly only, country to permit machine-generated patents to be granted may not turn out to be in Australia’s [our] best interests.

The main reasoning behind this “startup-mania” is that large companies are extremely good at developing incremental innovations, but not breakthrough innovations. There are many opinions, why it is so, like for example:

* Culture: Corporate culture usually motivates people to follow existing procedures, rather than being innovative.

One potential consequence of granting DABUS the right to invent and patent its own inventions is that this decision could open the floodgates of AI inventing. As more AI systems enter the patent scene, they theoretically could invent thousands of speculative ideas and have the right to sue anyone who accidentally stumbles across the same idea in the future. The heart of patent law is to encourage and protect innovation, not to patent thousands of speculative ideas just for inventing’s sake. Those against AI being given the rights of an inventor say this would only worsen the power of the tech elite – people who already have an inventing advantage and use that to profit off of smaller inventors who unintentionally infringe.

Further, if AI is given the right to inventorship, what would happen if that AI system infringed on another inventor’s patent? How could inventors be protected when AI faces no consequences for infringement?

The Al field already exhibits substantial innovation and disclosure under the current patent framework

The U.S. patent system exists to promote innovation and disclosure. Over the past decade, few fields have experienced a greater explosion in innovation than AI. There are new developments and advancements in AI technology all the time, as investment, research, and interest in the industry increased drastically over the past few years.18 AI development thrives everywhere, from universities and research institutions to startups and large companies. To remain globally competitive, the U.S. needs to keep investing in AI research and education.19

Today’s strategic environment features rapid technological change coupled with the increasing accessibility of cutting-edge technologies to more and more actors. These changes, and the threats they pose to U.S. national security, span presidential administrations and cross party lines. The 2018 National Defense Strategy, for example, notes that maintaining U.S. technological advantage requires significant changes across the “National Security Innovation Base” and calls on the Department of Defense (DoD) to “organize for innovation” and “out-innovate revisionist powers.”1 In his Interim National Security Guidance, President Joe Biden pledges to “sustain America’s innovation edge” and encourages “the culture of innovation required to address today’s complex challenges.”2 Over the past several years, DoD has attempted to adapt to this dynamic technological landscape by establishing organizations and concepts such as the Defense Innovation Unit, the Chief Digital and AI Office, Joint All-Domain Command and Control, Project Convergence, Project Overmatch, the Rapid Defense Experimentation Reserve fund, and a network of “tech bridges” focused on leveraging artificial intelligence, machine learning, and other advancements to develop sensor networks and shorten kill chains.3

The 2018 National Defense Strategy specifically mentions China’s military modernization, which has been so massive in scale and rapid in speed, that it threatens the U.S. military’s technological edge. Innovation, therefore, is a must. Since before the Spanish-American War, innovation has been a key “way” of U.S. military strategy. Out-innovating our adversaries has been a hallmark of the U.S. military—think stealth, high-bandwidth communications, space-based intelligence, surveillance, and reconnaissance, and precision-guided munitions. But just as sea power was taken for granted after the end of the Cold War, the Pentagon’s ability to innovate faster than its adversaries became a given—one that is no longer. Witness China’s antiship ballistic missiles, hypersonic weapons program, informationized warfare, and military space capabilities. The imperative to innovate is back.

The personal touch with which the author can stamp their work is the result of a complex intellectual process; a process that incorporates mostly the deconstruction of all the elements they receive, the conscious processing of ideas, images, sounds, emotions and senses and finally the composition of the above with a conscious choice, or with a conscious randomness.

Questions about the impact of technology on human creativity were raised three to four decades ago because of the so-called computer-generated works. First, the UK incorporated in its copyright law (Copyright, Designs and Patents Act 1988/CDPA) a provision for computer-generated works, i.e., works generated by computer in circumstances such that there is no human author of the work. As provided for by the UK law, as author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken. [15] Classifying a computer programme into the category of the author's tool became more widely accepted. [16] Simply put, a music software used to create new content represents an asset for the composer as does a camera for a photographer or a brush for a painter. It was a compromise option for integrating computer-generated works into the current legal system ascribing authorship to the individual who coordinates, controls, and possibly intervenes with the result generated by a computer programme. [17]

III. Do current patent laws and regulations regarding inventorship need to be revised to take into account inventions where an entity or entities other than a natural person contributed to the conception of an invention?

Q3: Do current patent laws and regulations regarding inventorship need to be revised to take into account inventions where an entity or entities other than a natural person contributed to the conception of an invention?

In a similar vein, there are some people in the field of intellectual property who are starting to ask questions about whether computers can be ‘creative’ or ‘inventive’ and, if so, whether it should be possible for a computer to be named as an inventor on a patent application – or, conversely, whether some humans should be disentitled from inventorship on the basis that their computers, rather than themselves, were the true inventors. One academic who has been making a name for himself in this emerging field of study is Professor of Law and Health Sciences at the University of Surrey, Ryan Abbott. Professor Abbott is the author of, among other works on the topic, ‘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), in which he argues that the law should embrace treating non-humans as inventors because this ‘would incentivize the creation of intellectual property by encouraging the development of creative computers.’

As I shall explain, however, I do not agree with Professor Abbott that computers can, or should, be regarded as inventors for the purpose of granting patents. Furthermore, while Abbott accepts claims that patents have already been granted on what he calls ‘computational inventions’, I firmly believe that a computer is yet to ‘invent’ anything. In my view, the researchers and technologists who claim otherwise have an interest in promoting a particular perspective, and in doing so they are subtly extending the definitions of ‘creation’ and ‘invention’ to encompass the contribution of their machines, to the detriment of the human operators who are responsible for providing the true creative input in the process.

When it comes to the specific field of hypothetical “AI-generated” inventions, the question that needs to be answered is whether this type of inventions is compatible with the theoretical foundation of the patent system described above. There are several reasons why this compatibility is not attainable from different perspectives. First, it can be stated that software developers would have non-economic incentives to justify the creation of computational computers.135 For instance, they would implement creative machines to enhance their reputation, or simply to satisfy a scientific curiosity and build up collaboration with other scientists.136 Therefore, from the perspective of creators, patents would not be needed to incentivise the development of this AI. Secondly, if patents are conferred for AI-generated inventions there could be an impairment of human innovation, since human creativity would be substituted and weakened by autonomous (not automated, for the difference see 2.3) algorithms.137 This leads in turn to the almost complete elimination of human talent and the easier availability of the inventive process to people who are not so skilled or full of resources.138 As the last relevant reason why AI inventions are not compatible with this justification of the patent system, although it is clear that innovation increases economic growth, if there is too much innovation, as it would happen in case of the fast development of always new “AI-generated” inventions, there would eventually be a lack of value for those inventions, since the consumers would fail to recognise real and true ground-breaking technologies.139 As can be observed, both the patent holder and the public at large would be deeply influenced by the patentability of AI-generated inventions, not in the sense of an unbalance between the different interests at stake (something that in the end could be accepted), but because there would be an impairment of all these interests. Therefore, from both perspectives, this incentive justification of the patent system does not seem compatible with creative machines.

What Are The Patenting Challenges When Using AI To Develop Drugs?

While AI is having a tremendous impact on drug development, a lack of clarity in patent law for AI-driven innovations means that companies may find it difficult to patent their inventions. For example, patent applications that use AI may face patent eligibility challenges as being directed to unpatentable subject matter including abstract ideas or laws of nature. It is also unclear how much disclosure in a patent application is necessary to satisfy the written description and enablement requirements, especially when drafting claims directed to antibodies (a complicated and unclear area on its own). These hurdles will each be discussed in subsequent articles.

In the case of the screwdriver or the drill, human direction and control are typically manifest and proximate. The human hand may be less apparent in the case of machines that are used to tool or assemble other machines. Over the past several decades, human intervention has become causally attenuated, so that the human prime mover may appear to have disappeared from the system. As mechanisms become more complex, human causation becomes more distant, but not necessarily less proximate--there is no obvious human agency directing the robots that assemble automobiles on a factory floor, yet we have no doubt that a human agency arranged, orchestrated, and initiated their activity. Human involvement in the chemical sciences is in some sense even more attenuated; human manipulation is not present or required in each reactive step. Rather, humans create the entropic conditions for reactions, and having initiated them, allow them to proceed. Biotechnological inventions encompass vast arrays of such reactions.

Patent inventorship and patent infringement may therefore be regarded as questions of causality--how far up the chain of causality must one look to find the human initiator? It may be that in some cases the human is sufficiently far removed that there is no inventor or infringer; if Bowman receives as a gift patented seeds that are caught by the wind and grow as sports in his field, perhaps he is no longer the proximate cause of making the claimed invention even if he is a cause in fact. Such matters of causality are old and familiar (if perhaps intractable) problems across the domains of law, spanning criminal culpability, tort liability, and a host of other areas. 67 Infringement causality is clearly undertheorized in patent law, as some commentators such as Amy Landers and Dmitry Karshtedt have begun to recognize. 68 Those cracks in patent causality have begun to show in recent decisions involving multiple actors and divided infringement, and will perhaps be salient in some future scenarios involving AIs.

The manifest human guidance behind a factory assembly robot leaves no reason to believe that a machine that assembles other machines is an inventor, or an infringer, but anything other than a tool of the inventing or infringing humans who employed it. Yet when it comes to machine learning systems, the objection that might be raised against this logic is that AIs have "emergent" qualities, that their output is not pre-determined or specified in advance in the way that the output of an automobile factory robot would be. The activity of the factory floor robot is programmed and known (if it is not, then presumably something has gone terribly wrong). AI systems, on the other hand, are portrayed as developing unexpected and undetermined outputs, at least within the parameters specified in their design. For example, pharmaceutical design AIs may generate molecular structures and species that are unforeseen by the humans who set them in motion.

But, as discussed before with reference to biotechnology, patent law is already well familiar with biological machines that engage "emergent" processes and that assemble other devices--perhaps not fully autonomously, but by means of "black box" processes that humans do not fully understand. A machine that generates unforeseen outputs by design is by no means disengaged from human innovation; to the contrary, it is entirely the tool of human innovation. The specifics of an AI's output, or that of a hybridoma, may be unforeseen, but the general expectation of some output, and some output within a particular field, was surely intended in the machine's deployment by humans.

The Leahy-Smith America Invents Act of 2011 (AIA) included several statutory amendments that facilitate corporate patenting while still retaining the notion of human inventorship. The most direct of these is elimination of inventorship as a condition of patentability. Prior to 2011, U.S. law stated clearly that a patent could be invalidated based upon the fact that a non-inventor was listed as the inventor.61 That provision was eliminated by the AIA. In parallel, the new law also eliminated the date of invention as relevant to patentability. Thus, the new law provides a shift from a first-to-invent system to a first-to-file system.62 Coupled with these changes are new regulatory provisions that allow corporate employers to directly seek patent protection. Previously, applications were filed on behalf of the inventors themselves except in rare circumstances.

The revised law continues to provide some tools to ensure that the listed inventors are the true inventors, but those tools are quite limited. For instance, the new law includes a limited 1-year grace period for the true inventor to file for patent protection in situations where someone else falsely claimed inventorship. 63 Alternatively, procedures also narrowly provide for a derivation proceeding, but again with a tight deadline of 1-year following either publication or patenting.64 But derivation can only be found if the allegedly derived invention is “the same or substantially the same invention,” and so excludes a situation where a “derived invention” serves as the base but then new elements are added to the invention. And, if an inventorship listing error is found, the revised law allows for correction of inventorship even if the original listing constituted fraud.65 And, under the new statute, a patent remains valid if inventorship can be corrected.66

The character of war continues to evolve, and there is no doubt that AI will significantly contribute to elements of that evolution for the foreseeable future. However, there are risks to an overestimation of the rate of technological change and the role of advanced tech in future victory. In spite of many exciting technological developments, history and data suggest that we are probably in an evolutionary rather than revolutionary period. Furthermore, at least since the beginning of the modern age technological overmatch has been neither necessary nor sufficient for victory, and is about as predictive as a coin toss.

In its understandable zeal for technological advantage, the military should remember enduring truths about warfare. The nuance of force employment is key to any battle, and an over-emphasis on technology risks allowing blind spots that our future adversaries will be sure to exploit. Though we should certainly strive for cutting edge technology, if that pursuit comes at the expense of force employment investments, we may be unknowingly entrusting future success to little more than a coin toss.

Much of the debate over military AI leadership and U.S. technological competition with China hinges on the assumption that there is a significant first-mover advantage when it comes to these technologies, meaning the first to develop them could reap substantial economic and military effects. However, fear of pronounced AI first-mover advantages instead reflects how AI is prone to “overhyping” where incredibly high expectations of capabilities surpass the reality of what is possible. Overhyping can obscure real progress and generate an inappropriate perception of an ​​​​​​AI arms race that “misrepresents the competition going on among countries.” Any possible first-mover advantage for AI would be unsustainable.

AI is a general-purpose, enabling technology not dissimilar to electricity. Moreover, the private sector drives its development, rather than the defense sector. While technologies that are singularly applicable to military contexts diffuse more slowly, those that are multi-use like AI have the added prodding of market incentives to speed up their spread.

Yet the history of technological revolutions counsels against alarmism. Extrapolating from current technological trends is problematic, both because technologies often do not live up to their promise, and because technologies often have countervailing or conditional effects that can temper their negative consequences. Thus, the fear that emerging technologies will necessarily cause sudden and spectacular changes to international politics should be treated with caution. There are at least two reasons to be circumspect.

First, very few technologies fundamentally reshape the dynamics of international conflict. Historically, most technological innovations have amounted to incremental advancements, and some have disappeared into irrelevance despite widespread hype about their promise. For example, the introduction of chemical weapons was widely expected to immediately change the nature of warfare and deterrence after the British army first used poison gas on the battlefield during World War I. Yet chemical weapons quickly turned out to be less practical, easier to counter, and less effective than conventional high-explosives in inflicting damage and disrupting enemy operations.6 Other technologies have become important only after advancements in other areas allowed them to reach their full potential: until armies developed tactics for effectively employing firearms, for instance, these weapons had little effect on the balance of power. And even when technologies do have significant strategic consequences, they often take decades to emerge, as the invention of airplanes and tanks illustrates. In short, it is easy to exaggerate the strategic effects of nascent technologies.7

How large are the transfers to the RPX-identified public PAEs relative to the size of the market in which they operate? 27 The firms identify the markets in which they operate in their SEC 10K filings. Though some of the firms have some patent assets in sectors such as medical devices, for the most part they tend to operate in high technology markets.28 We confirm this by consulting RPX reports on litigation activity. As Table 6 shows, those reports indicate the NPErelated lawsuits occurred in high technology manufacturing, software, and related service industries. We also consulted the FTC (2016: 57), which states that “… the vast majority of patents held by Study PAEs were Information and Communications Technology (ICT) patents: 88% were in the Computers & Communications or Other Electrical & Electronic technology categories, and more than 75% of the Study PAEs’ overall holdings were categorized as software-related patents.”

TABLE 6 AROUND HERE

Nonetheless, we think the focus on patent trolls obscures a more complex set of challenges confronting the patent system. In this Article, we make three points about the problems commonly associated with trolls. First, patent trolls are not a unitary phenomenon. We see at least three different troll business models developing, and those models have different effects on the patent system. Second, patent assertions by practicing entities can create just as many problems as assertions by patent trolls. The nature of many industries obscures some of the costs of those assertions, but that does not mean they are cost-free. In addition, practicing entities are increasingly engaging in "patent privateering," in which product-producing companies take on many of the attributes of [*2121] trolls. Put differently, while trolls exploit problems with the patent system, they are not the only ones that do so. Third, many of the problems associated with trolls are in fact problems that stem from the disaggregation of complementary patents (patents that cover technologies used together in the same products) into too many different hands. That in turn suggests that aggregators might be reducing, not worsening, these problems (though, as we will see, the overall effects are ambiguous) and that "patent privateers" that spin off patents in order for others to assert them might make things worse. For this reason, patent reformers and antitrust authorities should worry less about aggregation of patent rights and more about disaggregation of those rights, sometimes accomplished by spinning them out to others.

Understanding the economics of patent assertions by both trolls and practicing entities allows us to move beyond labels and the search for "bad actors" and to focus instead on aspects of the patent system itself that give rise to the problems and on specific, objectionable conduct in which both trolls and practicing entities sometimes engage. Patent trolls alone are not the problem; they are a symptom of larger problems with the patent system. Treating the symptom will not solve the problems. In a very real sense, critics have been missing the forest for the trolls. Exposing the larger problems allows us to contemplate changes in patent law that will actually tackle the underlying pathologies of the patent system and the abusive conduct they enable.

Second, critics gripe about the costs of troll lawsuits. 42 Bessen and Meurer (2014) estimated the sum of direct costs on defendants of nonpracticing entity (including troll) litigation to be approximately twenty-nine billion dollars. 43 Beyond direct lawsuit expenses, opponents argue that troll patent assertions also impose indirect costs, such as the disruption of business operations and loss of market share. 44 When examining the time period from January 2007 to October 2011, Bessen et al. (2011) found that defendants' total loss of wealth resulting from non-practicing entity litigation was eighty-three billion dollars per year. 45

[*43] While patent cases financially burden defendants, they also necessitate large capital investments from plaintiffs including trolls--a point that critics have failed to recognize. 46 For a median-value patent lawsuit between one and twenty-five million dollars, a plaintiff pays about two million dollars in litigation costs. 47 Unless a troll believes that the expected value of the case exceeds the expenses, it will not initiate the legal action. 48 Further, there is no empirical evidence that suggests troll litigation imposes higher costs on defendants than that of practicing entities. In fact, patent assertions by practicing entities are likely more burdensome for defendants due to the discovery needed to investigate these entities' products or services. 49

Trade secret protection offers several potential benefits for AI innovations:

AI technologies that are difficult to reverse engineer are particularly well-suited for trade secret protection. As discussed above, one major reason for pursuing patent protection is to prevent others from developing and marketing products using the protected technology. This is particularly the case where the technology could be readily reverse engineered by a competitor such that, absent patent protection, the innovation would be particularly vulnerable to copying. The drive to pursue patent protection is lessened where the underlying innovation is difficult to reverse engineer. By keeping such an AI technology as a trade secret, the inventor would be preventing others from having access to and potentially using the technology.