At the opening of this final session in the first cycle of the Global AI/IP Talks, a heartfelt acknowledgment was given to Professor Vishv Priya Kohli for her outstanding dedication and generous contributions throughout the entire series. Her consistent leadership and deep insight shaped each of the four encounters, creating a foundation for ongoing reflection and exploration at the intersection of Artificial Intelligence and Intellectual Property (AI/IP).
This fourth and concluding session focused on how AI can assist academics, students, and legal professionals in creating accurate and pedagogically valuable IP case law summaries. As with previous encounters, this session emphasized practical applications, methodological guidance, and the importance of human oversight when using AI tools in educational and professional IP contexts.
This event concluded the initial series of four AI/IP Talks organized by Fide and the European Intellectual Property Teachers’ Network (EIPTN), developed to help the IP community thoughtfully integrate AI into its educational practices.
Objectives
- Teach participants how to generate IP case law summaries using AI tools.
- Discuss workflow design, prompt engineering, and verification methods.
- Explore comparative applications of AI for summarizing decisions from various legal systems.
- Examine the limitations, risks, and ethical considerations of using AI-generated legal content.
Speakers
- Prof. Vishv Priya Kohli – Associate Professor, Copenhagen Business School. Member of the EIPTN Committee.
- Prof. Yolanda Bergel Sainz de Baranda – Associate Professor, Universidad Carlos III de Madrid. Coordinator of its IP Master Programme.
Moderator:
- Laurent Manderieux – Associate Professor of Intellectual Property Law at Bocconi University, Chair of the EIPTN, and Member of Fide’s Academic Council.
Key Discussions
1. Why Summarize Case Law Using AI?
Prof. Kohli outlined the pedagogical benefits of case law summaries:
- Facilitates student understanding of dense and fact-heavy decisions.
- Supports non-native English speakers in accessing complex judgments.
- Helps educators save time while designing instructional materials.
She emphasized that AI could identify essential elements (e.g. facts, issues, holdings, and doctrines) but warned that AI-generated outputs must be reviewed and edited by humans before use.
2. Tool Comparison: Notebook LM vs. ChatGPT
Two tools were showcased in the session:
- Notebook LM:
- Suitable for beginners.
- Generates summaries, briefing documents, and highlights key facts.
- Limited customization: editing must be done manually.
- ChatGPT (v4.5 recommended):
- Best for advanced users.
- Highly customizable via tailored prompts (e.g. asking for summaries suited for bachelor-level students, followed by reflective questions).
- Offers better control over structure, tone, and learning objectives.
Both tools were tested using real IP cases such as the Svensson v Retriever Sverige AB and the Lindt chocolate bunny trademark case, demonstrating how different prompts produce varying levels of nuance and detail.
3. Prompt Engineering: The Key to Succes
Prof. Kohli emphasized that well-crafted prompts are critical to obtaining reliable and pedagogically useful results. Examples included:
- “Summarize this case in 150 words for postgraduate students.”
- “Identify the ratio decidendi and generate three questions for class discussion.”
- “List key legal concepts in this decision and suggest three hypothetical scenarios for student analysis.”
She also explained how to test AI’s understanding of core legal concepts (e.g., «orbiter dicta» vs. «ratio decidendi») before relying on it for deeper interpretation.
4. Ethical Use and Verification
Prof. Kohli and Prof. Bergel repeatedly stressed:
- Never use AI summaries without cross-referencing the original case.
- Avoid uploading unpublished materials or third-party content without permission.
- AI tools should be considered sparring partners, not authorities.
Prof. Kohli presented a step-by-step workflow for ethical AI use, including validation of legal accuracy, completeness, terminology consistency, and purpose-driven adaptation (e.g., for group discussions or reflection exercises).
5. Interactive Dialogue: Use Cases and Limitations
- Prof. Bergel contributed practical queries and reflections throughout the session, reinforcing real-life classroom concerns.
- The discussion addressed:
- AI’s ability to work with non-English decisions (e.g. Spanish and Danish) — results are promising, but verification is essential.
- Whether future judicial decisions should be drafted in a more AI-accessible format — a proposal viewed as increasingly relevant.
- AI’s potential to distinguish between referenced decisions and the primary ruling — achievable with advanced models and clear prompting
Conclusion & Future Outlook
Moderator Laurent Manderieux praised the quality of the session and the impact of the full AI/IP cycle. He announced that resources for all four events (videos, reports, slides) are accessible on the GDE platform and encouraged continued engagement with the community.
He also marked this session as the culmination of the first AI/IP Talks series, but underscored that AI and IP is a living, evolving conversation, likely to continue with new formats and topics.
The initiative stands as a testament to international collaboration between the Fide Foundation via its Global Digital Encounters and the European IP Teacher’s Network – EIPTN, and to the commitment of educators like Prof. Vishv Priya Kohli, who have helped pioneer a practical, thoughtful path forward for AI in IP education.
