2026 PAW: Decision-Making Beyond Facts and Law at the 10th ICC European Conference

The week of the 23^rd of March marks the 10^thanniversary edition of Paris Arbitration Week (“PAW”). The 10^th edition opened with reflections on how the initiative has grown since its launch in 2017. What began with 17 partners now brings together 288 partners and approximately 40,000 registrations across the week’s events.

Opening the session, PAW co-presidents Diamana Diawara (ICC) and Catherine Schroeder (Schroeder Arbitration) reflected on what the next decade of PAW should look like. In a global context, where mechanisms for the peaceful resolution of disputes are increasingly challenged, PAW was purported as a forum for bringing together the international community and maintaining dialogue across jurisdictions.

PAW’s focus on diversity and inclusion was addressed in a conversation entitled “Diversity and Inclusion Beyond Words”, featuring Nania Owusu-Ankomah (Bentsi-Enchill, Letsa & Ankomah) and Coralie Lolliot (Sciences Po Paris), moderated by Yasmin Mohammad (Fortress Investment Group). The speakers addressed scholarship programmes, financial support, and the uncomfortable debate surrounding forms of positive discrimination. Lolliot noted that such measures can be controversial, particularly because individuals may fear being perceived as “tokens”. At the same time, they remain necessary to counter structural barriers within the profession. Owusu-Ankomah emphasised that the objective is not to lower standards but to widen access to opportunities and allow talent to become visible. Diversity initiatives therefore aim to broaden the road, rather than lower the bar. The opening session urged continued contributions to peaceful dispute resolution.

Why We Are All a Little Biased

The 10th ICC European Conference featured a conversation between Mihael Jeklic (King’s College London) and Alexander Fessas (ICC), exploring how cognitive biases shape decision-making.

One of the central questions was whether professional experience protects practitioners from cognitive bias. Jeklic explained that experience may help decision-makers adjust their expectations more accurately, particularly after observing a wide range of outcomes. However, it does not eliminate bias. Much of human reasoning operates unconsciously, meaning cognitive distortions remain inevitable.

The discussion focused particularly on myside bias—the tendency to favour information that confirms existing beliefs while discounting contradictory evidence. According to Jeklic, when individuals want a particular outcome to be true, they often struggle to evaluate evidence objectively. Predictions about how a tribunal or judge might decide may therefore be shaped by the same bias influencing the underlying assessment of the case. This becomes particularly visible in disputes where both parties predict a greater than 50 percent chance of success—an outcome that is statistically impossible. Arbitration, like litigation, may amplify this effect because the burden of persuasion rests largely on the parties themselves. As Jeklic noted, advocates are often more persuasive when they genuinely believe in their case.

Myside bias should not, however, be understood as unethical behaviour. Lawyers are not distorting their analysis in pursuit of fees or strategic advantage; rather, the bias reflects how the mind naturally tests hypotheses, often starting from an assumption that favours one side. Overconfidence can further intensify this dynamic. Jeklic described overconfidence as a “meta-belief” about the accuracy of one’s initial judgement. When individuals are highly confident in their first assessment, they become less likely to revisit or critically reassess it.

These biases may also affect settlement dynamics. While establishing direct causality is difficult, the correlation is clear: when both parties hold overly optimistic views of their chances of success, settlement becomes less likely. Jeklic illustrated this with a hypothetical example: one party values the case at 40 while the other expects 70. A settlement at 55 would benefit both sides, yet mutual overconfidence may prevent compromise.

The discussion also addressed possible ways to mitigate cognitive bias. Simply telling individuals not to be biased is ineffective. More promising techniques include asking lawyers to prepare arguments from the opposing party’s perspective, forcing them to reconsider initial assumptions. Jeklic also discussed experimental models comparing lawyers’ predictions with estimates generated by artificial intelligence (“AI”). When the cost of being significantly wrong increased—particularly where it risked losing a client—bias decreased significantly.

Ultimately, cognitive bias cannot be eliminated entirely. Some degree of bias may even be inherent to the role of advocacy itself. The greater concern arises on the side of decision-makers. Because cognitive bias often leaves no visible trace, it can operate unnoticed even in contexts that emphasise neutrality and independence. As Jeklic advised, whenever individuals have a personal stake—whether through advocacy or other forms of motivation—they should remain cautious about trusting their own judgement too blindly.

When Confidence Becomes Overconfidence

The first panel, moderated by Claudia Salomon (ICC), brought together Stephen Ruttle KC (Brick Court Chambers), Noradèle Radjai (Lalive), Gauthier Vannieuwenhuyse (TotalEnergies), and Todd Wetmore (Three Crowns) to examine how overconfidence affects advocacy, client advice and dispute strategy.

From counsel’s perspective, a certain degree of confidence is not only inevitable but often beneficial. Radjai observed that lawyers are expected to believe in their client’s case and that belief can shape arguments more effectively. The difficulty arises when confidence turns into overconfidence. When advocates become too invested in their own case theory, they risk overlooking weaknesses, misjudging the decision-maker’s perspective and therefore providing overly optimistic advice. Wetmore similarly emphasised that identifying weaknesses in a case is not harmful to advocacy; on the contrary, it strengthens it. Honest internal assessment allows counsel to test arguments and predict challenges from the opposing side.

From an in-house perspective, Vannieuwenhuyse highlighted the risks created by excessive optimism. Overconfidence can generate strategic blind spots and unrealistic expectations within an organisation. Early predictions of success often shape internal decision-making and can become difficult to revise later, undermining settlement opportunities and, when proven inaccurate, weakening internal trust in governance processes.

Ruttle, speaking from a mediator’s perspective, emphasised that mediation seeks to shift the focus away from percentages of success toward the human dimension of disputes. Rather than treating conflicts purely as probabilistic exercises, mediation aims to “turn litigants back into human beings”, encouraging parties to listen to one another and engage in constructive dialogue. Early involvement of mediators or facilitators, he suggested, may help parties reach more realistic discussions before positions become solid.

The panel also explored tools that might mitigate overconfidence. Radjai stressed the importance of maintaining internal objectivity while projecting external confidence. Practical techniques include assigning someone within the legal team to challenge the prevailing case theory or engaging neutral outsiders to test arguments. Wetmore also pointed to the growing role of AI in stress-testing legal strategies, for example by prompting systems to generate the opponent’s strongest arguments or simulate how a tribunal might reject a claim.

The panel agreed that while belief in one’s case is often necessary for persuasive argument, failing to critically examine that belief may ultimately weaken both strategy and credibility.

In line with the repeated references to AI throughout the discussions, the session also featured a brief presentation by Jean-Rémi de Maistre, Co-Founder and CEO of Jus Mundi introducing Jus AI, an AI tool designed to support arbitration research and case analysis.

How Tribunals Really Decide

The second panel, moderated by Erica Stein (Stein Arbitration), included Juan Fernández-Armesto (Armesto & Asociados), Hilary Heilbron KC (Brick Court Chambers), Stella Leptourgou (ICC), and Melanie van Leeuwen (Vanguard International Dispute Resolution) and explored how arbitral tribunals deliberate behind closed doors.

They first focused on tribunal composition. Leptourgou explained that appointing arbitrators carries significant responsibility for institutions such as the ICC. Beyond experience and expertise, attention is given to the balance within a three-member tribunal; appointing arbitrators with comparable levels of seniority can help preserve equality of voices. Cultural and professional diversity may also help counteract confirmation bias.

Van Leeuwen similarly emphasised that diversity within a tribunal—whether cultural, legal, or professional—can strengthen the deliberative process. When arbitrators approach a dispute from different perspectives, they are less likely to rely on shared assumptions and are instead forced to engage closely with the facts and arguments presented.

The panel also addressed the role of the tribunal chair in guiding deliberations. Heilbron stressed the need to establish a cooperative atmosphere from the outset. Practical steps, such as meeting co-arbitrators before hearings and encouraging less experienced members to question witnesses or counsel, can help ensure that all tribunal members feel comfortable contributing.

Attention then turned to dissenting opinions. Fernández-Armesto distinguished between good-faith dissents arising from genuine disagreements on law or evidence and strategic dissents driven by reputational considerations. While dissents may delay the issuance of awards, they can also strengthen decision-making by forcing tribunals to justify their reasoning more clearly. In some cases, a dissent may even persuade the other members of the tribunal and lead to a unanimous outcome.

From an institutional perspective, Leptourgou explained that the ICC typically becomes aware of tribunal dynamics only at the scrutiny stage. In most cases, dissenting opinions concern disagreements over factual assessments and are handled smoothly. Scrutiny may even improve the award by ensuring that the issues raised in a dissent are fully addressed in the majority’s reasoning. In more complex situations, however, the institution will intervene to ensure dissenting opinions do not undermine the award’s integrity or enforceability.

This panel confirmed that while cognitive biases cannot be entirely eliminated, careful tribunal composition, well-managed deliberation practices and institutional oversight can help maintain the quality and legitimacy of arbitral decision-making.

Following the Money

Moderated by Carole Malinvaud (Gide Loyrette Nouel), the final panel featured Sébastien Bonnard (Hughes Hubbard & Reed), Alina Dumitrașcu (PPC), Fabiano Robalinho Cavalcanti (Bermudes Advogados), and Ayse Yazır (Bench Walk Advisors) and examined how fees, funding and cost allocation shape the behaviour of parties and counsel.

From an in-house perspective, Dumitrașcu emphasised that general counsels are often required to provide early assessments of both the prospects of success and the expected costs of proceedings. In practice, this involves discussions not only with legal teams but also with senior management, including CEOs and CFOs, who must weigh litigation risks against broader business and regulatory considerations. Particularly in the energy sector, where disputes may involve state-owned entities or complex frameworks, these financial and strategic considerations often determine whether a dispute is pursued or settled.

The discussion then turned to how fee arrangements may influence litigation strategy. Bonnard noted that fee structures inevitably affect behaviour. Time-based billing offers flexibility and transparency but may raise concerns about cost discipline and open-ended proceedings. Fixed fees improve predictability and may encourage efficiency, though they often lead to negotiations over the scope of work. Phase-based budgeting was presented as a middle-ground solution, allowing parties to estimate costs at key milestones. At the other end of the spectrum, success-fee arrangements shift most of the financial risk to counsel and may affect incentives regarding the speed and conduct of proceedings.

Yazır provided the perspective of litigation funders, explaining that funders evaluate cases primarily through the lens of expected recovery and enforceability, often conducting detailed due diligence on both the legal merits and the parties involved. According to Yazır, settlement remains common in funded disputes.

Finally, Cavalcanti addressed the institutional perspective on costs. He noted that ICC arbitration rules grant tribunals broad discretion when allocating costs and emphasised the institution’s ongoing efforts to improve cost management. Mechanisms such as the potential reduction of arbitrators’ fees in cases of delay also reflect attempts to align financial incentives with procedural efficiency.

The Bigger Picture

Bringing all these panels together in a single day, the 10^th ICC European Conference took participants from the psychological instincts that shape decisions and strategies to the dynamics and financial incentives that frame outcomes, revealing how much more than facts and legal reasoning lies behind an award.

This post is part of Kluwer Arbitration Blog’s coverage of Paris Arbitration Week 2026