Working seminar between the Supreme Court of Singapore and the Cour de cassation

Chief Justice Sundaresh Menon opened the meeting by praising the dynamic cooperation between the two courts. First President Christophe Soulard then introduced the round tables, recalling the challenges posed by artificial intelligence in the judicial world. He presented AI as both a risk and an opportunity to make justice more accessible and predictable. In this sense, he examined the new disputes that have arisen (such as the Thaler v Comptroller-General of Patents, Designs and Tramarks case in the United Kingdom). He also questioned the use of AI by legal professionals, raising the risk of a disembodied justice and the advent of de facto jurisprudence, recalling the work carried out on the initiative of Chief Justice Donal O'Donnell within the framework of the network of the Presidents of the Supreme Judicial Courts of the European Union.

At the first round table on the use of new technologies and AI by the supreme courts, President Sandrine Zientara, Director of the Documentation, Research and Reporting Department, described the Open Data project for court decisions piloted by the Court of Cassation and the artificial intelligence initiatives developed to pseudonymise decisions made public. She also mentioned the possibility of using AI to detect emerging and serial litigation, in connection with the Court's project to set up an Observatory of Judicial Litigation. The aim here is to encourage dialogue between judges, both horizontally and vertically, and to enable the Court to have a precise view of the state of play in the handling of a new issue by all the substantive courts before ruling on it itself. Judge Aedit Abdullah of the Singapore Court said that from 2019 a judge of the Court had been assigned to reflect on technological innovation and an Office of Transformation and Innovation had been opened at the Court. It has developed the various algorithmic technologies that have been put in place, such as an engine that can determine civil liability (in the event of an accident) and the amount of compensation for damage. The Singapore Court has also introduced a system for authenticating decisions, with an electronic signature and QR code archiving of decisions. In an even more innovative move, Judge Aedit Abdullah explained that since 2023, in addition to e-appeals, an "asynchronous hearing" has been in place, enabling non-simultaneous exchanges between judges and counsel for the parties for the preparation of cases. He concluded his speech by mentioning the prospects for collaboration with legal tech start-ups, with the aim of building a chatbot that could answer litigants' questions about how to act in a given situation. The example given was that of a litigant asking a question about his noisy neighbour.

The second round table looked at positive law and AI. From the point of view of criminal law, Judge Isabelle Goanvic, member of the Criminal Chamber, presented the existing normative framework at European and national level. She emphasised that the malleability of criminal fault in French law was such as to allow the characterisation of common law offences in the context of AI. In procedural terms, she also highlighted the challenges, particularly in terms of personal data protection in the context of police files, for example.

Judge referee Caroline Azar, member of the First Civil Chamber and special adviser to the First President of the Cour de cassation, looked at AI protection litigation in the field of intellectual property and patent law. She then asked about compensation for any damage caused by an AI or an autonomous robot. Finally, Caroline Azar concluded her speech by discussing the relevance of the regulations currently being drawn up at European level.

In reply, Mr Justice Goh Yihan distinguished the different faculties of AI in order to consider the possible responsibilities involved. For example, he developed the case of Quoine v. B2C2 to explain that in the context of self-learning AI, the discrimination that could result from poor learning was attributable to the designers, who had not anticipated the bias adopted by the machine.

The exchanges between the two lectures on this highly topical subject were fruitful. After describing the regulatory landscape in each State and the initiatives put in place in each Court during the seminar, it was agreed to continue this work on targeted topics of substantive law and judicial organisation. These high-level operational discussions are an important part of the Court's international strategy, which is supported by the First President and promotes comparative law and the inspiration of good practice.