News
Read our latest news, stories and articles


Three Minds, One Award: Tribunal Dynamics explored at the ASA Fall Conference 2026 

Nhu-Hoang Tran Thang, Astute Dispute Resolution

What goes on behind closed doors when three arbitrators come together as a tribunal? On the afternoon of 25 September 2026, Catherine-Anne Kunz (LALIVE) and Anna Masser (Manner | Masser) convened arbitration practitioners at the ASA Fall Conference in Bern to look behind the curtains of arbitral tribunals and explore the dynamics that drive them. Under the theme “Tribunal Dynamics”, the conference traced the progression of a case across three panels, from the constitution of the tribunal to the drafting of the award. 

By inviting the audience to shape the discussion through the submission of questions in advance, the conference offered a practical and candid examination of the inner workings of arbitral tribunals, a topic rarely addressed so openly. 

Panel 1 – Pre-Hearing Phase: Tribunal Constitution and Case Management

Moderated by Rahul Donde (Rahul Donde Dispute Resolution), the first panel featured Paolo Michele Patocchi (Patocchi & Marzolini), Inka Hanefeld (HANEFELD) and Angeline Welsh KC (Essex Court Chambers). 

The panel commenced with an examination of protocols for the selection of the presiding arbitrator. List procedures, pursuant to which each party and its co-arbitrator contribute names to a joint list that the parties subsequently rank, were described as effective when co-arbitrators act jointly at every stage and when the protocol clearly defines what constitutes failure and what consequences follow. Experiences varied considerably, however, as to the length of the process, from presidents appointed within two months to constitution phases exceeding a year. One panelist observed a trend away from prescriptive lists towards informal consultation between the co-arbitrators and the parties. The panel concurred that selecting a president who is acceptable to both parties and constitutes a good fit for the tribunal is among the most important responsibilities of a co-arbitrator. 

Interviews of prospective arbitrators were addressed candidly. The panelists expressed no objection to interviews being recorded, and agreed that the impermissible boundary is any discussion of the merits of the case. Abstract questions concerning experience and inclination are appropriate, but candidates should never provide assurances as to how they would decide. Interviews were described as the most efficient means to assess a candidate’s command of the language of the arbitration, and it was suggested that counsel share the topics of discussion in advance of the interview. At the same time, panelists expressed skepticism that a brief conversation reveals much about how a candidate will conduct themselves within a tribunal. Speaking with those who have previously sat with the candidate is often more informative. 

Asked whether linguistic and regional backgrounds are material within Switzerland, one panelist distinguished cultural affinity, which approaches bias and is of limited relevance, from procedural culture, which is pertinent. Practitioners trained in the German tradition may, for instance, be more inclined to raise the prospect of settlement facilitation with the parties. Even so, the panelist cautioned against categorizing practitioners by region.  

On case management, the president’s draft Procedural Order No. 1 is the natural starting point, open to modifications from counsel that are reasoned and specific. The question of whether to address AI in the first procedural order divided the room. Some panelists had observed sophisticated parties and tribunals remain silent on the topic, given the rapidity with which the technology evolves. Others had adopted straightforward provisions confirming that the parties remain responsible for the accuracy and confidentiality of their work and that the tribunal will not use AI for decision-making. A quick poll indicated that only a minority of the audience wished AI to be addressed at the outset. One panelist questioned the added value of such provisions, which restate principles that are self-evident, while another considered them a useful way to set out the rules of the road in an emerging area. 

Panel 2 – Hearing Phase: Preparation and Dynamics 

The second panel, moderated by Filip Boras (Baker McKenzie), brought together Gabrielle Nater-Bass (Homburger), Domitille Baizeau (LALIVE) and Raëd Fathallah (Bredin Prat). 

The panelists provided a rare insight into how arbitrators prepare for a hearing, and the various approaches in this regard. One approach commences as soon as the statement of claim is received, with an issue matrix that evolves throughout the proceedings, from a neutral chronology to the legal requirements of each prayer for relief and the evidence supporting or contradicting each factual proposition. Another concentrates on the period immediately preceding the hearing and begins with the exhibits, read chronologically, in order to understand the narrative before examining the parties’ pleadings. A third combines both, allocating one to two weeks before the hearing to construct one’s own chronology, which frequently reveals issues that neither party has fully addressed, notably the applicable law. The common thread was that preparation does not aim to decide the case in advance, but to master the record sufficiently to identify what must be tested at the hearing. 

The moderator then enquired whether opening statements continue to serve a purpose before such well-prepared tribunals. One panelist firmly rejected any suggestion of dispensing with them. Opening statements are the juncture at which counsel bring the case together and focus the tribunal on the key issues, and in this panelist’s experience they have invariably enhanced the tribunal’s understanding of the dispute. 

The use of AI in hearing preparation generated a nuanced discussion. Some panelists declined to use AI-generated summaries, citing insufficient confidence in their accuracy and a risk of bias — for example when tools accord disproportionate weight to repeated arguments — while employing AI to review drafts or search the record. Others considered that the quality of the output depends on the prompts, and that AI used in a neutral, symmetrical and controlled manner, for instance to test two expert reports against identical questions, can assist in preparing the examination of experts. One panelist emphasized that AI should never be asked to assess the credibility of a witness. 

Views also diverged on drafting portions of the award before the hearing and on pre-hearing working sessions among arbitrators. Maintaining the procedural history and the summary of the parties’ positions up to date can provide a neutral working document for the tribunal, while others prefer to draft nothing beyond the procedural aspects to avoid any risk of prejudgment. Pre-hearing working sessions were described as valuable provided they remain focused on identifying open questions rather than deciding them, while a meeting the evening before the hearing offers a lighter alternative. 

On the conduct of the hearing, engaged co-arbitrators were regarded as preferable to silent ones. One panelist nonetheless considered it inappropriate to interrupt a cross-examination, and suggested that sensitive questions be posed through the president. On inequality of arms, the panelists cautioned both against assuming the role of advocate for the weaker party and against dismissing a poorly presented case prematurely. 

Panel 3 – Post-Hearing Phase: Deliberations and Drafting of the Award

Moderated by Nina Lauber-Thommesen (Independent Arbitrator), the final panel featured Yoshimi Ohara (Nagashima Ohno & Tsunematsu), John V H Pierce (Latham & Watkins) and Philippe Pinsolle (Quinn Emanuel Urquhart & Sullivan).  

The panelists first examined how to prepare and commence deliberations. The sequence in which issues are addressed was identified as determinative, as it may shape the outcome. Decision trees and high-level lists of issues were recommended as useful instruments, as was the practice of initiating informal deliberations during the hearing itself, so that issues surface while the parties can still address them. Whether the chair should speak first depends on the circumstances: in straightforward matters it may facilitate consensus, while in cases where strong views are held it may be prudent to let the co-arbitrators to speak first. 

Posing questions to the parties after the hearing was considered appropriate, and sometimes necessary, where outcome-determinative issues have not been addressed, particularly where the tribunal intends to rely on a legal ground that the parties have not debated, which in certain jurisdictions, such as France, may expose the award to annulment. The difficulty lies in framing questions with sufficient precision without appearing to prejudge the outcome, and it was emphasized that deciding on a point the parties never argued constitutes the greater danger. An intervention from the floor, referring to the English High Court of Justice’s decision in Federal Republic of Nigeria v P&ID ([2023] EWHC 2638 (Comm)), raised the question of whether tribunals should intervene when a party is inadequately represented. The panel distinguished cases of fraud, where the tribunal has a duty to inquire, from instances of ordinary deficient advocacy, and noted divergent sensitivities between common law and civil law traditions. 

On disagreements within the tribunal, unanimity remains the objective. Dissenting opinions were regarded as justified only in rare instances of genuine and material disagreement, although several panelists observed that the mere possibility of a dissent frequently strengthens the reasoning of the award. The panelists were also reluctant to trade points within the tribunal simply to avoid a dissent—for instance by allowing a party to prevail on the merits in exchange for losing on a significant portion of quantum or on costs. The Japanese perspective added a comparative dimension: Article 63 of the JCAA Commercial Arbitration Rules prohibits arbitrators from disclosing their dissenting opinions in any manner. One panelist explained that this rule was introduced in response to past instances in which dissents had been employed to assist the appointing party, while expressing the hope that the JCAA would reconsider it. 

The role of the co-arbitrator was described as that of a constructive partner, ensuring that no fact or argument is overlooked and, in particular, that the arguments of the appointing party are properly considered, with appropriate timing and in genuine dialogue with the other members of the Tribunal. An exchange with the floor clarified the distinction between this role, which pertains to the process, and impartiality, which governs the decision. The panel also recommended fixing deliberation dates at the earliest opportunity, ideally immediately following the hearing. Finally, the president should in principle hold the pen for the sake of consistency, and rendering an award “appeal-proof” begins with a thorough knowledge of the law of the seat. 

* * * 

The conference concluded with Closing Remarks by ASA President Andrea Meier (Walder Wyss), who recalled that every tribunal is different because it brings together three individuals, and that this human dynamic is precisely where the impact of AI may reach its limits.