IP and AI – for a Balanced, Non-Protectionist Stance

Journal article

1 online resource (page 1017–1018)

The development of EU legislation regarding IP is largely a history of a protectionist approach, which tends to expand the area of exclusive erga omnes rights, broaden their scope, and favor an ‘easy’ access thereto: ‘expansionist’ as Hanns Ullrich would label it. It objectively favors the interests of IPR holders (typically firms) for profit maximization, while stifling both dynamic competition and social access to information and culture. A new chapter of this story is apparently in progress. It concerns AI: precisely the IP status of the works/results that AI produces – the output – and the data that robots use according to AI’s (technical) ‘mode of production’ – the input. As to the output, European law (in the substantive synthesis of legislation and administrative practice) seems so far inclined towards a rather protectionist stance by generally admitting machine-produced works for IP protection. Put more clearly, it seemingly draws no consequence from the material ‘duality’ of AI technology (albeit evoked by its very definition in Art. 3 of the AI Act, ‘Varying levels of autonomy’): aspect one is predominantly ‘servient’ to detailed human instructions to achieve a certain goal/result; and aspect two is ‘generative’, pre-dominantly self-developing upon basic instructions to achieve a given result. This ‘generalist’ approach can be observed in the field of patents and copyright alike. In particular, as concerns the objective requirements of the patentability of software (AI is software, what else?), the EPO’s 2024 Guidelines stand on the traditional consolidated approach based on the assessment of a ‘further technical effect’ – a feature equally visible in AI aspect ‘one’ and AI aspect ‘two’ (Part G, II-3.1-3.3).

Crossref

GRUR International ; 73, 11, 2024, 2632-8550.

[Oxford, England] : Oxford University Press (OUP), 2024.

English

https://academic.oup.com/grurint/article/72/3/231/6998505

External Online Journals > GRUR