Categories
Artificial Intelligence

On AI and Inventorship: Copyright & Patents

On the 29th of September, Professor Desmond Osaretin Oriakhogba delivered his highly anticipated presentation on AI, Inventorship, Patents and Copyright. Professor Oriakhogba traversed a range of topics and the full recording may be found here. This follow-up serves to provide a brief recap of the discussion. 

The presentation commenced with the definition of AI and how it relates to patents. Professor Oriakhogba distinguished between two groupings of AI: Narrow and General AI. Narrow AI refers to AI systems that can perform individual or specific problems within clearly defined parameters such as language translations, chatbots and image recognition. General AI refers to artificial intelligence capable of behaving and surpassing human intelligence. It is the former, rather than the latter, that was central to his discussion. 

Firstly, Professor Oriakhogba considered AI systems as computer programs and as such, they qualify as copyright work due to the protection extended to computer programs. Further, Professor Oriakhogba notes that, in the context of South Africa, the Copyright Act recognises authorship by humans and corporations but our courts have not yet had the opportunity to decide whether AI may be authors under the Copyright Act. Moving to different jurisdictions, the Indian Copyright Office named an AI as a co-author on a painting app developed by an AI system. And in China, there are two cases of relevance: Shenzhen Tencent v Shanghai Yingxun and Beijing Feilin Law Firm v Baidu. In the Tencent case, an article was written autonomously by AI but the court placed authorship on the developers of the AI. Similarly, in the Baidu case, the court stated that Chinese copyright law only recognises works created by humans and juristic persons.

He then moved the crux of the discussion, whether AI-generated works are patentable. Section 25 of the Patent Act outlines the requirements for an invention to be patentable, these are invention must be : novel, involve an inventive step and capable of trade, industrial and agricultural application but excludes computer programs. The pertinent question then becomes, whether this exclusion applies to AI-generated or AI-assisted inventions? If this exclusion does not apply to AI-generated works, can AI be named an inventor in an application for patent in respect of such invention?

As noted in a previous blog post, DABUS is a project aimed at challenging settled foundational principles of inventorship in patent law. They have filed applications to have DABUS named the inventor in various jurisdictions and failed in all, bar South Africa. The CIPC recently granted a patent to DABUS in respect of an application filed by Dr Stephen Thaler where the AI system was named as the inventor. Professor Oriakhogba notes that this grant is wrong for three reasons: (i) It runs counter to the human-centric nature and foundation of Patent Applications, (ii) It uncritically accepted a PCT application (PCT is a convention that permits international  patent filling procedure) under the unjustifiable belief that PCT applications are not subject to formal examinations at the national stage as to the nature of the inventor and the invention and (iii) an applicant who is not the inventor must show proof of his title, which is usually an assignment by the actual inventor. For a more comprehensive discussion, see: DO Oriakhogba ‘What If DABUS Came to Africa? Visiting AI Inventorship and Ownership of Patent from the Nigerian Perspective’ (2021) 42(2) Business Law Review 89 & ‘DABUS gains territory in South Africa and Australia: revisiting the AI-inventorship question’ (2021) 9 South African Intellectual Property Law Journal 87.

See you all at the next talk!

Leave a Reply

Your email address will not be published.