Assessing Intellectual Property Relevant Similarities In Images Through Algorithmic Decision Systems

The question of whether there are similarities between two given objects is a central one in the field of intellectual property (IP) rights. The answer is an extremely complex one, better left to IP offices and judges—through administrative and legal procedures, respectively—, who do not have proper analysis tools.

However, algorithmic decision systems (ADS) have been developed, mainly by private companies, and they are now increasingly used to enforce IP rights (monitoring counterfeit products online, content filtering) and for registration by IP offices.

In order to limit biases and protect values of public interest associated with IP, the development of these tools should ideally be carried out under the supervision of independent experts. With this in mind, the project intends to develop an open, supervised, and transparent model to analyse IP similarities (IPSAM), specifically for 2D images.

This interdisciplinary project will have legal practitioners (JurisLab) and engineers (LISA) examine the methodological, technical, legal, and ethical challenges involved in developing such tools, in order to provide a critical study of the technological solutions currently available.

The project’s result will enable original contributions to the debates surrounding algorithmic regulation in general and in the field of IP in particular.

Coordination : Julien Cabay, Centre de Droit privé, Faculté de Droit et de Criminologie


Created on September 4, 2020