[en] The European patent system was designed around a paradigm of human inventorship. This paper will analyse in depth and from a de lege ferenda perspective the rather general arguments against and in favour of a possible designation of artificial intelligence (AI) systems as inventors. For the sake of a more concrete discussion, it will also outline a potential reform of the European patent system to implement AI inventorship and allocate the right to the European patent for such inventions by default to the machine’s operator. In the process, it will highlight the major specific issues associated with a reform that acknowledges AI inventorship and touch upon possible alternative approaches to addressing the growing autonomy of machines within the R&D process. The study must not be understood as a call for a reform to recognise AI systems as inventors but rather as a manner of laying the foundations for a more concrete, critical and fruitful discussion on non-human inventorship and its alternatives. The analysis will show that the more general, highly conceptional reservations advanced in the current discussion against AI inventorship are somewhat unfounded, e.g. the alleged break with the functions of the current patent system or the alleged need to endow AI with legal personality. More convincing arguments against a reform that allows for the designation of AI systems as inventor might instead relate to the specific difficulties associated with such reform.
Disciplines :
Droit, criminologie & sciences politiques: Multidisciplinaire, généralités & autres
Auteur, co-auteur :
STIERLE, Martin ; University of Luxembourg > Faculty of Law, Economics and Finance (FDEF) > Department of Law (DL)
Co-auteurs externes :
no
Langue du document :
Anglais
Titre :
A De Lege Ferenda Perspective on Artificial Intelligence Systems Designated as Inventors in the European Patent System.
Date de publication/diffusion :
2021
Titre du périodique :
Gewerblicher Rechtsschutz und Urheberrecht. Internationaler Teil
The European Patent Convention was signed on 5 October 1973 and entered into effect on 7 October 1977. The conceptual foundations, however, are based on the structures of the national patent systems which were mainly developed in the 19th and at the beginning of the 20th century.
In this paper the term autonomous is used to describe an enhanced level of machines independent of human input. For a technical distinction of automation and autonomy see Daria Kim, AI Generated Inventions : Time to Get the Record Straight? [2020] GRUR International 446 f.
cf Lewis Mumford, Technics and Human Development (Harcourt, Brace & World 1967) 191 (defining a machine, more or less in accord with the classic definition of Franz Reuleaux, as a combination of resistant parts, each specialized in function, operating under human control, to utilize energy and to perform work [. . .] ).
Martin Stierle, Artificial Intelligence Designated as Inventor An Analysis of the Recent EPO Case Law [2020] GRUR International 918, 923.
See for the US: USPTO, In re application of 16/524, 350. See for the UK: UKIPO, Decision of 4 December 2019, BL O/741/19, on GB1816909.4 and GB1818161.0 as well as the appeal Thaler v The Comptroller-General of Patents, Designs and Trade Marks [2020] EWHC 2412 (Pat).
See EP18275163.6 and EP18275174.3.
EPO, Grounds for decisions of 27 January 2020 on EP18275163.6 (para 19 ff) and EP18275174.3 (para 20 ff).
Besides art 81, r 19(1) EPC as one ground for the decision, the EPO argued that the indications by the applicant would not meet the requirements of art 81, 60(1) EPC (EPO, Grounds for decisions of 27 January 2020 on EP18275163.6 (para 30 ff) and EP18275174.3 (para 31 ff)). This finding is surprising in the light of art 60(3) EPC (cf critically Stierle (n 4) 922).
cf UKIPO, Decision of 4 December 2019, BL O/741/19, on GB1816909.4 and GB1818161.0, para 29 and CIPA, Patenting Inventions created using an AI system, A CIPA Discussion Paper https://www.cipa.org.uk/_resources/assets/attachment/full/0/260456. pdf accessed 27 July 2020; Andreas Engel, Can a Patent Be Granted for an AI-Generated Invention? [2020] GRUR International, sub I.1. (forthcoming).
Another major technology that might boost the autonomy of AI systems is unsupervised training. For this method see for example Nils J Nilsson, The Quest for Artificial Intelligence: A History of Ideas and Achievements 513 ff https://ai.stanford.edu/_nilsson/QAI/qai.pdf accessed 27 July 2020, and Yann LeCun, Yoshua Bengio and Geoffrey Hinton, Deep learning [2015] Nature 521, 436, 442 with further references.
Allen Newell and Herbert A Simon, Computer Science as Empirical Inquiry: Symbols and Search [1976] CACM 113, 116.
Vincent C Mü ller and Nick Bostrom, Future progress in artificial intelligence: A survey of expert opinion in Vincent C Mü ller (ed), Fundamental Issues of Artificial Intelligence (Springer 2016) 553
Martin Ford, Architects of Intelligence: The Truth about AI from the People Building it (Packt Publishing 2018) 528
David Silver and others, Mastering the game of Go with deep neural networks and tree search [2016] Nature 529, 484.
cf Erica Fraser, Computers as inventors legal and policy implications of artificial intelligence on patent law (2016) 13(3) SCRIPTed 305, 323 with further references.
Peter Stone and others, Artificial Intelligence and Life in 2030, One Hundred Year Study on Artificial Intelligence: Report of the 2015-2016 Study Panel, Stanford University https://ai100.stanford.edu/2016-report accessed 27 July 2020
Ekkehard Ernst, Rossana Merola and Daniel Samaan, The economics of artificial intelligence: Implications for the future of work (ILO 2018) https://www.ilo.org/wcmsp5/groups/public/ dgreports/ cabinet/docu ments/publication/wcms_647306.pdf accessed 27 July 2020.
See Herbert Zech, Life Sciences and Intellectual Property: Technology Law Put to the Test (2015) 7 IPJ 1, 3 (describing different functions of technology law, the first being influencing the creation of technology).
David L Poole and Alan K Mackworth, Artificial Intelligence: Foundations of Computational Agents (2nd edn, Cambridge University Press 2017)
Josef Drexl and others, Technical Aspects of Artificial Intelligence: An Understanding from an Intellectual Property Law Perspective (2019) Max Planck Institute for Innovation and Competition Research Paper No 19-13 https://papers.ssrn.com/sol3/papers.cfm?abstract_id 3465577 accessed 27 July 2020.
Ryan Abbott, I Think, Therefore I Invent: Creative Computers and the Future of Patent Law (2016) 57 B.C. L. Rev. 1079, 1079 ff
Peter Blok, The inventor s new tool: Artificial intelligence how does it fit in the European patent system? (2017) 39 E.I.P.R. 69, 70
Shlomit Yanisky Ravid and Xiaoqiong Liu, When Artificial Intelligence Systems Produce Inventions: An Alternative Model for Patent Law at the 3A Era (2018) 39 Cardozo L. Rev. 2215, 2219
Nick Li and Tzeyi Koay, Artificial intelligence and inventorship: An Australian perspective (2020) 15 JIPLP 399, 400
Noam Shemtov, A study on inventorship in inventions involving AI activity (EPO, February 2019) 22 http:// documents.epo.org/projects/babylon/eponet.nsf/0/3918F57B010A3540C 125841900280653/$File/Concept_of_Inventorship_in_Inventions_involv ing_AI_Activity_en.pdf accessed 27 July 2020
Yann Ménière and Heli Pihlajamaa, Kü nstliche Intelligenz in der Praxis des EPA [2019] GRUR 332, 335
Lea Tochtermann, Immaterialgü terrechtlicher Schutz von KI de lege ferenda in Markus Kaulartz and Tom Braegelmann (eds), Rechtshandbuch Artificial Intelligence und Machine Learning (Vahlen 2020) 7.3 para 38
https://www.wipo.int/export/sites/www/about-ip/en/artificial_intelligence/conversation_ip_ai/pdf/igo_epo.pdf accessed 27 July 2020.
See Abbott (n 18) 1082 ff (2016) for alleged examples of such computers generating patentable results.