26 mars 2024
13h - 14h - CET
Gratuit, mais inscription en ligne requise pour recevoir l'invitation
Prof. Julien Cabay and Thomas Vandamme will make a presentation on AI & TM Search : “Legally Blind? Findings on Transparency and Explainability through a Study of AI-Powered Trademark Search Engines”, followed by a discussion with the audience.
♦ Attendance is free but registration online is required to receive the invitation.
The question whether two objects are similar in a relevant manner is core to intellectual property law (IP). The answer is extremely complex and entirely left to IP Offices and judges, respectively in the frame of administrative and judicial proceedings, without any proper analytic tools. Yet, algorithmic decision systems (ADS) are currently being developed and used by private companies for the purposes of IP enforcement (monitoring infringing goods online, filtering out content) and registration by IP Offices, outside of public scrutiny.
In the framework of our research project IPSAM (Intellectual Property Similarities Assessment Model), we subjected several publicly available trademark search engines to close scrutiny.
In the course of our study, we encountered several methodological and practical difficulties that in turn, evidenced normative issues related to transparency and explainability. Should we compute and process possibly inconsistent case law? How can we identify legal biases in technical solutions? Is it possible to study technologies despite trade secret protection? What methodology can we develop for evidencing overfitting? Should we limit transparency to AI qualify as ‘high-risk’?
In the framework of our research project IPSAM (Intellectual Property Similarities Assessment Model), we subjected several publicly available trademark search engines to close scrutiny.
In the course of our study, we encountered several methodological and practical difficulties that in turn, evidenced normative issues related to transparency and explainability. Should we compute and process possibly inconsistent case law? How can we identify legal biases in technical solutions? Is it possible to study technologies despite trade secret protection? What methodology can we develop for evidencing overfitting? Should we limit transparency to AI qualify as ‘high-risk’?
And the critical question: are we legally blind?
Julien Cabay is full time Professor and Director of JurisLab at Université Libre de Bruxelles (ULB), as well as Associate Professor at Université de Liège. Besides, he is currently affiliated with KU Leuven, CEIPI (University of Strasbourg), Digital Law Center (University of Geneva), and a member of the Belgian Council for Intellectual Property. In the past, he has been Research Fellow and PostDoc Researcher at National Fund for Scientific Research, Visiting Research Fellow at Columbia University in the City of New York and Global Policy Fellow at Instituto de Tecnologia e Sociedade do Rio de Janeiro. His main fields of expertise are Copyright Law; AI & IP; Fundamental Rights & IP; IP & Open Science.
Thomas Vandamme is an engineer from ULB (specialty in Electronics and Telecommunications) currently in his final year of PhD. His doctoral thesis, promoted by Pr. Olivier Debeir (Engineering Faculty, ULB) and Pr. Julien Cabay (Law Faculty, ULB) is at the crossroads of their respective domains: the application of Deep Learning methods to the problem of Intellectual Property Similarity Assessment. The recent developments of his thesis have drawn him towards the evaluation of those systems, as a first step towards transparency and explainability
