Search the archive.

Mimi Afshar, Winter 2022, “Artificial Intelligence and Inv tificial Intelligence and Inventorship – Does the P orship – Does the Patent Inv atent Inventor Have to be Human?” Hastings Science and Technology Law Journal, https://repository.uchastings.edu/cgi/viewcontent.cgi?article=1110&context=hastings_science_technology_law_journal

Some scholars have called for the amendment of patent laws to eliminate the statutory requirement of listing the inventor.96 However, this proposal would deprive human inventors of the benefit of inventorship because one of the perks for inventing is the moral recognition of the inventor among his or her peers.97 While this argument has merit, it may still be plausible to make an exception and eliminate this requirement for those inventions where AI is the only inventor.

Erica Fraser, December 2016, “Computers as Inventors – Legal and Policy Implications of Artificial Intelligence on Patent Law,” https://script-ed.org/article/computers-as-inventors-legal-and-policy-implications-of-artificial-intelligence-on-patent-law/

The final option is to eliminate the statutory requirement to identify an inventor altogether. This option would be the simplest in the cases where there is no individual human inventor who can reasonably lay claim to an invention, such as where the invention is the result of a large team effort, or of a computer whose process of invention is not purposeful or directed and is without human oversight or control. However, eliminating this requirement would obviate certain justifications for the patent system, such as the fair reward and moral benefit of recognition. This could have tangible impacts on scientists and engineers who gain professional credibility and even monetary benefits based on their status as a named inventor on patents. Further, since the requirement is based to a large degree on a natural understanding of fairness, eliminating it could undermine public support for the system as a whole. In the current context and with the prospect of autonomous computer-generated inventions, the optimal balance between perceptions of morality and fairness and the impersonal realities of inventorship is likely to maintain the option of identifying one or more human inventors only where such a person can be reasonably identified. This would maintain the social and financial benefits of attribution where it is warranted, while permitting the patenting of computational inventions without making radical legal changes.

A second option is to conclude that the use of AI should not render unpatentable an otherwise patentable invention. This would preserve the innovation-promoting incentives that the patent system was designed to foster. But it would lead to complex questions of inventorship. As the International Association for the Protection of Intellectual Property stated in a 2020 resolution, “given the ability of AI systems to ‘learn,’ traditional notions of inventorship may be challenged.” Under this second option, it seems reasonable to attribute inventorship to the natural persons behind the AI, thereby avoiding the many problems that would arise if AI systems were deemed inventors. Of course, this attribution process would raise its own substantial complexities given the varying roles people play in designing, programming, training, and evaluating the outputs of AI systems. And it would require rethinking the current framework under which, as the U.S. PTO explains in a widely referenced but non-binding guidance document, the “threshold question in determining inventorship is who conceived the invention.”

The increasing use of sophisticated Artificial Intelligence (AI) technologies in conceptualizing new inventions has led to an intense international debate on patent law implications in relation to innovations generated by or with the assistance of AI technologies.1 In particular, can AI generate inventions autonomously, ie without a significant human contribution? If so, are such AI-generated inventions protected under current patent laws, and should they be? Who could and should be recognised as an inventor and owner of such patents? Stakeholders, policy makers, courts and academics around the world have been actively examining these issues.2 In parallel, a similar discussion is happening in other areas of IP law, such as copyright, with no international consensus on whether creations generated with assistance from AI could and should be afforded copyright protection.3 The importance of these issues is unquestionable: the decision to award or not award patent or copyright law protection for AI-generated outputs is likely to affect the future utilisation of AI technologies in innovation and creative processes as well as the further expansion of AI technology sectors around the globe.4 The recent UK Supreme Court decision in Thaler v Comptroller of Patents, Designs and Trademarks,5 which is a part of an international Artificial Inventor project led by Dr Stephen Thaler and his legal team,6 has provided an ultimate clarification on some of the questions regarding the patentability of AI-generated inventions, at least in the UK.7 Namely, the Supreme Court unanimously concluded that, under the Patents Act 1977, an AI system cannot be listed as an inventor in the patent application, and that the owner of the AI system that arguably generated the invention is not eligible to apply for a patent in this invention or to be granted a patent over it.8 This outcome is consistent with the outcomes of most other patent applications that Dr Thaler submitted in other jurisdictions and reinforces an emerging international consensus that AI cannot be an inventor for the purposes of current patent law.9 The next part of this note provides the factual background of Thaler v Comptroller of Patents, Designs and Trademarks and briefly reviews the decisions of the UK Intellectual Property Office (UKIPO), the High Court, the Court of Appeal and the UK Supreme Court. The following three parts focus on three main issues examined by the Supreme Court: i) whether AI can be listed as an inventor; ii) whether the AI owner can apply for and be granted a patent for an invention autonomously generated by a machine; and iii) whether the UKIPO can and should assess the truthfulness of the inventorship statement. The note concludes that the human-centric approach to inventorship, as adopted by the Supreme Court, is well-reasoned and maintains the internal consistency of the Patents Act 1977. While it might disincentivise the use and development of AI technologies with autonomous inventive possibilities, it helps ensure active human involvement in technologically-supported innovation and supports human ingenuity more generally. Further, despite the decision focussing on current patent law only, it will make it more difficult to advocate for future patent laws that include non-human inventors. In addition, the decision helps reveal challenges to and gaps in patent law that have emerged as a result of AI technologies, such as determining the ownership of AI-assisted inventions and ensuring the truthfulness of inventorship claims when AI is used in the inventive process. Such issues will require legal intervention and redress by law makers heading into the future as the use of AI is expected to proliferate.

As discussed earlier, a company’s return on research investment is in the form of patent ownership and control, not simply having a name on a patent. Patent ownership allows companies to bolster their market position by gaining market exclusivity or to block competitors’ entry into the market, to sell additional products through bundling arrangements or licensing, or simply to own the patent for a rainy day. These motivations are completely independent of the arrangement of inventor bragging or credit rights and focus solely on ownership rights. From the perspective of the company, then, whether a human, an AI, or no one at all is listed on the patent as an initial matter makes no difference. Ownership will accrue to the company contractually, and they make their money either way. This implies that employee compensation doesn’t change depending on who is on the patent. If the patent is equally valuable to the company in any scenario, they equally value its creator in any scenario. From an economic perspective, then, engineers should receive the same pay from their employer, the same prize bonus for a patent, the same increase in likelihood for internal promotion, regardless of who is on the patent listing as an initial matter.

As regards Issue 1, the first question, i.e., whether an AI can be considered as inventor, should be discarded taking into account the considerations regarding legal personality of AI. There are too many ethical implications as regards equating a machine to a natural or legal persona in the sense that they can be holders of rights. Opening that debate for the time being, even for the purposes of patent law, may result in expanding that debate to other areas of law that are clearly unprepared, such as tort law and liability, tax law or labour rights.

In the case of whether a human inventor should be name, the vast majority of national and regional legislations (e.g. section 81 of the European Patent Convention) require naming a human author in the patent application. We suggest two different options:

A large majority of American voters are skeptical of the argument that the U.S. should race ahead to build ever more powerful artificial intelligence, unconstrained by domestic regulations, in an effort to compete with China, according to new polling shared exclusively with TIME. The findings indicate that American voters disagree with a common narrative levied by the tech industry, in which CEOs and lobbyists have repeatedly argued the U.S. must tread carefully with AI regulation in order to not hand the advantage to their geopolitical rival. And they reveal a startling level of bipartisan consensus on AI policy, with both Republicans and Democrats in support of the government placing some limits on AI development in favor of safety and national security. According to the poll, 75% of Democrats and 75% of Republicans believe that “taking a careful controlled approach” to AI—by preventing the release of tools that terrorists and foreign adversaries could use against the U.S.—is preferable to “moving forward on AI as fast as possible to be the first country to get extremely powerful AI.” A majority of voters support more stringent security practices at AI companies, and are worried about the risk of China stealing their most powerful models, the poll shows. The poll was carried out in late June by the AI Policy Institute (AIPI), a U.S. nonprofit that advocates for “a more cautious path” in AI development. The findings show that 50% of voters believe the U.S. should use its advantage in the AI race to prevent any country from building a powerful AI system, by enforcing “safety restrictions and aggressive testing requirements.” That’s compared to just 23% who believe the U.S. should try to build powerful AI as fast as possible to outpace China and achieve a decisive advantage over Beijing. The polling also suggests that voters may be broadly skeptical of “open-source” AI, or the view that tech companies should be allowed to release the source code of their powerful AI models. Some technologists argue that open-source AI encourages innovation and reduces the monopoly power of the biggest tech companies. But others say it is a recipe for danger as AI systems grow more powerful and unpredictable. “What I perceive from the polling is that stopping AI development is not seen as an option,” says Daniel Colson, the executive director of the AIPI. “But giving industry free rein is also seen as risky. And so there’s the desire for some third way. And when we present that in the polling—that third path, mitigated AI development with guardrails—is the one that people overwhelmingly want.” The survey also shows that 63% of American voters think it should be illegal to export powerful AI models to potential U.S. adversaries like China, including 73% of Republicans and 59% of Democrats. Just 14% of voters disagree. A sample of 1,040 Americans was interviewed for the survey, which was representative by education levels, gender, race, and the political parties for whom respondents cast their votes in the 2020 presidential election. The margin of error given for the results is 3.4% in both directions. So far there has been no comprehensive AI regulation in the U.S., with the White House encouraging different government agencies to regulate the technology themselves where it falls under their existing remits. That strategy appears to have been put in jeopardy, however, by a recent Supreme Court ruling that limits the ability of federal agencies to apply broad-brushstroke rules set by Congress to specific, or new, circumstances. “Congress is so slow to act that there's a lot of interest in being able to delegate authorities to existing agencies or a new agency, to increase the responsiveness of government” when it comes to AI policy, Colson says. “This [ruling] definitely makes that harder.” Even if federal AI legislation seems unlikely any time soon, let alone before the 2024 election, recent polling by the AIPI and others suggests that voters aren’t as polarized on AI as they are on other issues facing the nation. Earlier polling by the AIPI found that 75% of Democrats and 80% of Republicans believe that U.S. AI policy should seek to prevent AI from quickly reaching superhuman capabilities. The polls also showed that 83% of Americans believe AI could accidentally cause a catastrophic event, and that 82% prefer slowing down AI development to account for that risk, compared to just 8% who would like to see it accelerated.

Samuel Scholz, "A Siri-Ous Societal Issue: Should Autonomous Artificial Intelligence Receive Patent or Copyright Protection?," Cybaris®: Vol. 11 : Iss. 1 , Article 3. https://open.mitchellhamline.edu/cybaris/vol11/iss1/33.

To counteract the issue of increasing the time needed to rule on a patent, the USPTO will need to hire additional examiners to review applications. Examiners are educated and well paid, as the USPTO requires examiners receive a four-year bachelor’s degree and their average salary is $124,675.63.189 Hiring many new examiners with well above average annual salaries will create an additional cost on society. Additionally, examiners would likely be limited in how much time they may allocate to reviewing any application. As the prior art expands, and the time limits for examiners remains the same, the likelihood of mistakenly granting a patent that should have been rejected for lack of inventive step or anticipation increases. This would likely shift a portion of the burden of reviewing the prior art to the judicial system, as courts will need to determine whether a patent is valid after it has been granted.190 Shifting this burden to the courts not only creates additional stress on the judicial system, but it imposes a massive financial burden on inventors. The average cost of receiving a software patent is about $16,000, 191 while the average cost of litigating a successful software patent through final disposition that earns between one to twentyfive million dollars is $2.8 million.192 This litigation cost of one hundred and seventy-five times the original cost of receiving a patent motivates inventors to be certain that their idea is patentable before filing an application. Inventors rely on the USPTO to perform an excellent review of the prior art before granting a patent to avoid future litigation costs.

What happens if the judges rule in favor of AI? For his part, Abbott said the existing U.S. patent law could be interpreted in such a way to allow for AI as inventor. “We argue it’s critical to interpret the patent act consistently with its purpose — to promote innovation,” Abbott said. “That’s why the law should allow protection for AI-generated inventions.” “The entity that conceived of the invention is at issue in this case,” he added. “DABUS is the factual inventor of these products. We argue the factual inventor must be the legal inventor.” If the ruling allows patents on AI-generated inventions, DABUS would not be the patent holder. Thaler would hold that patent, as a machine can’t own property, Abbott said. “It’s the same way, if you own a 3D printer that makes an object for you, you own the rights to the object itself,” he said. Should the judges rule in favor of AI, “you’ll see a lot more patent filings,” Vinti said. “This could open up Pandora’s box if programmers can claim inventorship for AI.” The claim to the invention is a big deal because, recently, courts have seen many arguments about “proper inventorship,” Vinti said. “If you name the improper inventor, that patent is dead.” Anytime programmers identify themselves as an inventor, he added, “I’ll dig in and see if it was created by AI, and then you could fight over, did you really invent this or did the computer and, if so, what happens?”

Enrico Bonadio, Luke McDonagh, and Plamen Dinev, “Artificial Intelligence as Inventor: Exploring the Consequences for Patent Law,” Intellectual Property Quarterly, 03-22-2021, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3798767

In addition to the legal considerations above, there are also sound policy reasons to deny large volumes of nonsensical computer-generated text prior art status, including the need to discourage the fabrication of knowledge aimed at preventing competitors from obtaining patents in that specific area. Such disclosures could be considered as being artificially constructed in bad faith by entities which are exclusively interested in preventing the acquisition of patents by those who actually invest time and resources in research and development. Hattenbach and Glucoft further argue that publishing large amounts of nonsensical claims may impinge on the fundamental aims of the patent system and ultimately have a negative effect on social welfare: Publishing masses of nonsense achieves the opposite of what these requirements seek to accomplish—it dilutes the set of actual public knowledge, burying genuinely useful information and leaving society worse off.48

Di Minardi, The grim fate that could be ‘worse than extinction,’ BBC, October 15, 2020, https://www.bbc.com/future/article/20201014-totalitarian-world-in-chains-artificial-intelligence

Researchers at the Center on Long-Term Risk, a non-profit research institute in London, have expanded upon x-risks with the even-more-chilling prospect of suffering risks. These “s-risks” are defined as “suffering on an astronomical scale, vastly exceeding all suffering that has existed on Earth so far.” In these scenarios, life continues for billions of people, but the quality is so low and the outlook so bleak that dying out would be preferable. In short: a future with negative value is worse than one with no value at all.

Tim Dornis, “Artificial Intelligence and Innovation: The End of Patent Law as We Know It,” Yale Journal of Law & Technology, Fall,

Yet the effects of “artificial inventions” on the contents and, ultimately, on the theoretical understanding of the state of the art may prove detrimental—at least in the long run. Of course, we might expect the emergence of new types of inventions and discoveries, maybe so elaborate and unexpected and beyond human understanding that the future will become one big bang of knowledge.177 But we might also expect the opposite. After all, if the state of the art is to be increasingly shaped by artificial actors and artificial inventions, issues of content and quality will come to the surface. This does not primarily concern whether inventive AI acts illegally or unethically. It is the simple fact that inventions are artificial that is problematic, especially since our theoretical understanding of the state of the art will inevitably deteriorate over time.

Natasha Rao, February 17 2021, “Are AI-Generated Inventions the Future?” International Trademark Association, https://www.inta.org/perspectives/features/are-ai-generated-inventions-the-future/

Even if IP law could adapt to recognize a non-human inventor, this solves only half of the problem. An AI does not have a legal personality, it is precluded in any event from being party to any employment agreement, from owning any IP rights, and from being able to transfer any such IP right to a third party (such as its owner).

Twelfth Judicial Circuit, “Guardianship Basics,” https://www.jud12.flcourts.org/About/Divisions/Probate-Guardianship/Guardianship-Basics

Guardianship is the process designed to protect and exercise the legal rights of individuals whose functional limitations prevent them from being able to make their own decisions and have not made plans for this time in their lives. People who need guardianship may have dementia, Alzheimer’s disease, a developmental disability, chronic illness or other such conditions that generally cause functional limitations. Before a guardianship is established, it must be determined that the alleged incapacitated person (AIP) lacks capacity. Guardianship should be the last resort.

Roman V. Yampolskiy, "Human Indignity: From Legal AI Personhood to Selfish Memes," arXiv:1810.02724, 10-2-2018, https://arxiv.org/abs/1810.02724

5. Human Safety

Roman V. Yampolskiy, 10-5-2018, "Could an artificial intelligence be considered a person under the law?" Conversation, https://theconversation.com/could-an-artificial-intelligence-be-considered-a-person-under-the-law-102865

Reducing human status

Jonathan Sutin, "State v. Montoya, 2017 NMCA 33," New Mexico Court of Appeals, CourtListener, 2-24-2017, https://www.courtlistener.com/opinion/4383977/state-v-montoya/

Personhood

Frank DeCosta, August 16 2022, “Drawing a ‘Bright Line’ Rule over AI Inventorship,” Finnegan, https://www.finnegan.com/en/insights/articles/drawing-a-bright-line-rule-over-ai-inventorship.html

The ruling touches on the key issue of whether an AI should own property. Under US law, patent ownership initially vests with the inventor. A patent is a property right that can be assigned, sold, pledged as collateral, licensed, enforced in court, etc. If AI were permitted to be an inventor it would, through the grant of a patent, have a property right.  This raises a policy question: do we want AI to have property rights (or any type of rights)?

David G. Henry, "Subject: Opposition to AI-Inventor Patents," Email to aipartnership, 9-24-2019, https://www.uspto.gov/sites/default/files/documents/David-Henry-01_RFC-84-FR-44889.pdf

In my opinion, patents should not be granted on the basis of artificial intelligence “inventors”. As I have taught my law students for many years, the patent system (as well as other intellectual property protections) represent a carve‐out from our otherwise free enterprise system. The latter serving the purpose of providing ever better and more affordable goods and services through the process of competition.

Adam Adler, "Oops, We Just Took Our First Real Step Toward the Machine Uprising," The Escapist, 8-14-2022, https://www.escapistmagazine.com/robot-ai-inventor-us-court-patent-law-machine-uprising/

Second, the decision avoided a robot-rights problem that would have arisen had the court reached the opposite conclusion. The question at issue was whether an AI system could be listed as an inventor on a patent application. Even if the answer had been yes, the AI system would not have had any ownership rights in the resultant patent — those rights would go to Stephen Thaler, the person who created the AI system and filed the patent application. (Under the patent laws, the owner of a patent and the inventor of a patent can be different.)