| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.26 MB | Adobe PDF |
Authors
Abstract(s)
Este estudo aborda a interação da Inteligência Artificial (IA) no processo inventivo e subsequente elegibilidade para patenteabilidade das criações daquela proveniente. A maioria dos regimes jurídicos dos diversos países e organizações internacionais não reconhece nem a pertinência nem a atribuição de autoria de uma invenção a uma IA. Dada a complexidade das atuais IAs e, ainda, a dificuldade latente na compreensão e na reprodutibilidade da invenção gerada com intervenção destas, a patenteabilidade fica comprometida (e, bem assim, todo o processo de reconhecimento e reivindicação de direitos associados). A tendência internacional, nessa vertente formal da patenteabilidade, incide sobretudo na capacidade e necessidade de se possuir o conhecimento, suficientemente claro e aceitável, do modo de funcionamento da IA. De como, e porquê, esta última, chegou a esse resultado nessa mesma invenção. O Direito alcança a sua verdadeira utilidade no presente e futuro de uma Sociedade, e esta não estagna à espera de clarificações exatas que tarde, ou, até mesmo nunca, ocorrem. Poderá a solução passar pela determinação de uma dimensão aceitável de dúvida a valorar? Equaciona-se a ponderação sobre o facto de, a existir a intervenção da IA, esta relevar na aceitação da atribuição de autoria num processo de patenteabilidade de uma invenção. Quer afastando, pura e simplesmente, a titulação da invenção nos termos legais em vigor (apontando para um regime sui generis); quer desconsiderando essa contribuição e outorgando a autoria ao inventor natural ou, ainda; quer atribuindo parte da autoria à (s) pessoa (s) humana (s) mantendo, contudo, uma obrigação de indicação que essa invenção teve uma intervenção relevante de IA para ser escrutinada de forma diferenciada, tendo em conta essa mesma componente.
Esta situação configura uma preocupação a nível mundial pois acaba por afetar e delongar a inovação tecnológica e, por conseguinte, fragiliza a própria Economia.
Almeja-se, com o presente estudo, contribuir para esta temática, através do levantamento das principais posições e orientações emergentes na área, culminando na apresentação das nossas próprias perspetivas, com o propósito de mitigar o referido dilema.
This study deals with the interaction of Artificial Intelligence (AI) in the inventive process and the subsequent eligibility for patentability of creations arising from it. Most of the legal systems in the various countries and international organisations do not recognise either the relevance or the attribution of authorship of an invention to an AI. Given the complexity of today's AIs and the latent difficulty in understanding and reproducing the invention generated with their intervention, patentability is compromised (as well as the whole process of recognising and claiming associated rights). The international trend in this formal aspect of patentability focuses on the ability and need to have sufficiently clear and acceptable knowledge of how AI works. How, and why, the latter arrived at that result in that very invention. Law achieves its true usefulness in the present and future of a Society, and Society does not stagnate waiting for exact clarifications that are late, or even never, in coming. Could the solution be to determine an acceptable level of doubt to be valued? We are considering whether, if AI intervenes, it will be relevant in accepting the attribution of authorship in an invention's patentability process. Either by simply removing the title to the invention under the legal terms in force (pointing to a sui generis regime); or by disregarding this contribution and awarding authorship to the natural inventor; or by awarding part of the authorship to the human person(s) while maintaining an obligation to indicate that this invention had a relevant AI intervention in order to be scrutinised in a differentiated way, taking this same component into account. This situation is a worldwide concern because it ends up affecting and delaying technological innovation and, consequently, weakening the economy itself. The aim of this study is to contribute to this issue by surveying the main positions and emerging guidelines in the area, culminating in the presentation of our own perspectives, with the purpose of mitigating this dilemma.
This study deals with the interaction of Artificial Intelligence (AI) in the inventive process and the subsequent eligibility for patentability of creations arising from it. Most of the legal systems in the various countries and international organisations do not recognise either the relevance or the attribution of authorship of an invention to an AI. Given the complexity of today's AIs and the latent difficulty in understanding and reproducing the invention generated with their intervention, patentability is compromised (as well as the whole process of recognising and claiming associated rights). The international trend in this formal aspect of patentability focuses on the ability and need to have sufficiently clear and acceptable knowledge of how AI works. How, and why, the latter arrived at that result in that very invention. Law achieves its true usefulness in the present and future of a Society, and Society does not stagnate waiting for exact clarifications that are late, or even never, in coming. Could the solution be to determine an acceptable level of doubt to be valued? We are considering whether, if AI intervenes, it will be relevant in accepting the attribution of authorship in an invention's patentability process. Either by simply removing the title to the invention under the legal terms in force (pointing to a sui generis regime); or by disregarding this contribution and awarding authorship to the natural inventor; or by awarding part of the authorship to the human person(s) while maintaining an obligation to indicate that this invention had a relevant AI intervention in order to be scrutinised in a differentiated way, taking this same component into account. This situation is a worldwide concern because it ends up affecting and delaying technological innovation and, consequently, weakening the economy itself. The aim of this study is to contribute to this issue by surveying the main positions and emerging guidelines in the area, culminating in the presentation of our own perspectives, with the purpose of mitigating this dilemma.
Description
Keywords
Propriedade Industrial Propriedade Intelectual Patenteabilidade Inteligência Artificial Invenção Autoria Black-box
