THE AUTHOR:
Madhav Singal, India-Qualified Advocate & LL.M. Candidate – Stockholm University
On April 15, 2025, during the 31st Willem C. Vis International Commercial Arbitration Moot in Vienna, Jus Mundi hosted an event on the intersection of Artificial Intelligence (“AI”) and arbitration. Titled “AI & Arbitration: From Classroom to Courtroom”, the event brought together leading voices in arbitration and legal technology, including Marina Stanisavljevic, Counsel, Schoenherr; Cláudio Finkelstein, Professor, Pontifical Catholic University of São Paulo; Alexandre Vagenheim, VP of Global Legal Data, Jus Mundi; Paula Eiben-Działoszyńska, Vis Moot Coach, Sciences Po TADS Teaching Assistant & Legal Consultant, World Bank; and moderated by Melis Mani, Legal Counsel, Strabag Group.
‘Jus AI’
The event kicked off with a demonstration of ‘Jus AI’, a new AI model being developed by Jus Mundi over the past two years. Alexandre showcased the tool’s ability to respond to complex legal prompts with detailed, source-linked answers, accompanied by a list of those sources provided at the end of each answer. Jus AI can summarize party submissions and translate documents into multiple languages. This allows for an analysis of an arbitral award in Arabic to be first provided in English, and then further translated, for example, into Japanese!
Notably, Jus AI can even identify gaps and weaknesses in the uploaded documents, such as hearing transcripts, by leveraging the Jus Mundi database, all while ensuring data confidentiality and data protection regimes like the GDPR. For practitioners and students, ‘Jus AI’ holds promise in streamlining research, document analysis, and even hearing preparation.
He noted challenges such as ‘AI hallucination’ and ‘bias’ within AI and its implications in the legal scenario. To combat this issue, the Jus AI model is fine-tuned using multilingual datasets, which significantly diminishes the possibility of language-related bias. He further explained that the model uses Retrieval-Augmented Generation (“RAG”), which essentially confines the AI to a specific dataset. This means the model generates responses solely based on the provided document; hence, each answer is accompanied by footnotes and source references for individual paragraphs. The model is primarily designed for counsels rather than judges or arbitrators, as deploying AI in the administration of justice and human decision-making is a highly sensitive matter that demands great caution.
Bridging the Practical Gap Between Arbitration Professionals and AI
Artificial intelligence is valuable in streamlining workflows, but should not be over-relied upon. Furthering the discussion, Marina provided a practitioner’s lens on AI’s role in arbitration. Recalling her early days in the profession, of sorting through paper bundles to prepare cost submissions, she described how AI now assists in managing voluminous documentation, from email chains to fact-finding reports.
AI enhances efficiency by cutting down physical labour but cannot replace intellectual labour and strategic judgment unique to the legal profession. Viewing AI as a tool for delegation, she pointed out concerns about young professionals delegating their work to AI, which raises issues of competence and originality. She further drew attention to the “human touch” in client interactions, noting that AI-generated responses can feel robotic. In a highly adversarial arbitration environment, where advocacy resembles a high-stakes tennis match, human intuition and judgment remain irreplaceable.
The Young Generations’ Perspective on AI
As a frequent user of AI, Paula highlighted its myriad benefits while cautioning users that AI should not be used as a ‘copy-paste’ tool. She emphasized the importance of understanding how and when to use AI tools responsibly, stressing that AI must not become a shortcut for thinking. Legal professionals must verify, contextualize, and critically assess AI outputs.
The use of AI in legal education, she noted, presents opportunities for workshops on ethical usage, data privacy, and skills-based training. Paula emphasized the diversity in learning styles among law students. Some are voracious readers, while others prefer practical applications. AI, she suggested, can support both ends of this spectrum, especially in mastering industry-specific knowledge and simplifying complex tasks, such as research or due diligence.
AI in Legal Academia: Changing the Way We Think?
Professor Finkelstein reflected on the transformative impact of AI in legal academia. Drawing from decades of experience, he observed a fundamental shift in how students think, write, and engage with legal problems. While academia has yet to formally standardise AI usage, he mentioned that many institutions still lack clear protocols. However, various tools for detecting AI-generated assignments are emerging.
Cláudio acknowledged the inevitability of AI in education but raised concerns around “outsourcing thinking.” He noted that AI can mirror individual reasoning patterns and perform tasks at an unprecedented scale, raising fears about its long-term effects on legal cognition and labour markets.
Cláudio’s remarks highlighted a key pedagogical question: at what point in the process of legal drafting does real thinking occur? As AI tools accelerate outputs, educators must remain focused on fostering analytical skills and ethical awareness among future lawyers.
Conclusion: Learning the Law and Learning to Work with AI
A resonant theme throughout the discussion was that AI is not replacing lawyers; rather, it is reshaping how we learn, think, and practice law. The next generation of arbitration professionals must not only master the law but also understand how to work with AI tools responsibly, ethically, and strategically. As a student of international commercial arbitration, this event reaffirmed the dual responsibility of embracing technological innovation while preserving the human elements of justice, i.e. critical thinking, empathy, and advocacy. In the classroom and the courtroom alike, AI is here to stay. The challenge ahead lies in learning how to use it wisely.
ABOUT THE AUTHOR
Madhav Singal is a mechanical engineer turned lawyer. He is enrolled in the LL.M. program in International Commercial Arbitration at Stockholm University. He has close to three years of litigation experience and has previously worked with the legal team at Reliance Industries Limited.
*The views and opinions expressed by authors are theirs and do not necessarily reflect those of their organizations, employers, or Daily Jus, Jus Mundi, or Jus Connect.
