Preprint
Article

This version is not peer-reviewed.

Is Artificial Intelligence Patentable in Europe? A Brief Overview According to EPO Jurisprudence

Submitted:

13 August 2026

Posted:

14 August 2026

You are already at the latest version

Abstract
This paper examines the patentability of artificial intelligence (AI)-related inventions under the European Patent Convention (EPC), with particular attention to the framework developed by the European Patent Office (EPO) for mathematical methods and computer-implemented inventions.AI-related inventions remain subject to the ordinary patentability requirements of novelty, inventive step and industrial applicability, but the analysis requires three distinct inquiries. Under Article 52 EPC, an AI or machine-learning model claimed merely as an abstract mathematical method may be excluded, whereas a claim involving technical means may possess technical character as a whole. Under Article 56 EPC, AI-related features can support inventive step only insofar as they contribute to a technical solution to a technical problem.Under Article 83 EPC, the application must disclose the model, training process, relevant data characteristics and other implementation details to the extent necessary for the skilled person to reproduce the claimed technical effect without undue burden.EPO decisions T 0161/18 and T 1669/21 illustrate that inadequate disclosure of a neural network or computational model may result in insufficiency under Article 83 EPC and, depending on the circumstances, may also prevent reliance on the alleged technical effect for inventive step. The paper also presents an exploratory Espacenet search and a broader EP full-text patent-landscape analysis using the EPO Technology Intelligence Platform (TIP), illustrating the growth of AI-related European patenting and the effect of prosecution lag on observed grant rates.
Keywords: 
;  ;  ;  
Subject: 
Social Sciences  -   Law
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

Subscribe

© 2026 MDPI (Basel, Switzerland) unless otherwise stated

Accessibility

Disclaimer

Terms of Use

Privacy Policy

Privacy Settings