Herman Croux chaired a session on the protection of Artificial intelligence (AI) at the IBA Annual Conference in Seoul on 23 September 2019. As AI is gaining popularity and business value, the providers of AI solutions are looking for protection of that value as well as for their other tangible assets. AI developments can contain patentable inventions. As an alternative to or in addition to a (public) patent, certain algorithms or know-how are not disclosed, and trade secret protection is pursued for these elements. Sonia Cooper of Microsoft and Hyon Jong Kim of the Korean telecom provider SKT gave concrete examples of patents, such as query endpointing based on lip detection. Dallas attorney Elizaveta Dolghih illustrated how companies put trade secret protection in place, and how they sometimes fail to do so in practice. Ultimately it are the data that constitute much of the AI value, which lead to a discussion on open data models. Presentations can be found at https://www.ibanet.org/