Abstract
The paper examines how these jurisdictions define inventorship, allocate responsibility, and regulate the interaction between human and machine creativity within the current intellectual property protection system. Whether Artificial Intelligence (AI) systems can qualify as inventors under national and international patent law has become as one of the most contested issues in contemporary intellectual property doctrine. The findings indicate that WIPO advocates for a neutral “AI-assisted inventorship” model, emphasising global harmonisation, whereas the EPO follows a human-centred approach that allows only natural persons to be considered inventors. This article argues that the core regulatory challenge is not recognising AI as an inventor, but allocating legal responsibility for AI-assisted inventive outputs. Uzbekistan’s patent law reserves inventorship exclusively for natural persons, thereby excluding AI from independent recognition. Recent AI governance instruments in Uzbekistan signal incremental alignment with WIPO standards (Presidential Resolution RP-358, 2024). These instruments do not alter Uzbekistan’s foundational patent doctrine. The study concludes that AI need not be recognised as an inventor to comply with WIPO and EPO standards. Instead, three mechanisms can close the gap without any legislative overhaul. These include mandatory disclosures of AI involvement in patent applications, mediated attribution rules that allocate responsibility among developers and applicants; and deepened institutional cooperation with the EPO. Together they offer Uzbekistan a path to meaningful alignment with international patent standards. The article further argues that the patentability of AI-generated inventions implicates questions of attribution justice, access equity, and accountability fairness that reform frameworks must address alongside doctrinal harmonisation.