Instructed Late, Judged Early: What the Ciarb 2026 report on expert witnesses asks of experts, tribunals and counsel

In this paper, Max Beckett and Catherine Jago from CJH Experts review the Ciarb Report on Party-Appointed Expert Witnesses in International Arbitration (the Ciarb Report)

The Ciarb Report was published in July 2026 based on 419 survey responses. Those responses, which came from arbitrators, experts, external counsel and in house counsel, were analysed by four regional working groups. The main finding is simple. Expert evidence works best when it is set up early, the technical approach agreed, and the process managed actively by the tribunal.  Every group will find something uncomfortable in the results, including experts. Broadly speaking, however, we believe it is an important step towards making the process more efficient, effective and professional.

In brief

For readers short of time, our five main points are these:

• First, every group agrees that the tribunal should read the first expert reports as soon as they are served and send the experts one list of questions.  Over three quarters of every group supports this.  In our experience, it is rarely done.

• Second, experts are instructed too late and then given too little time.  The Ciarb Report calls late instruction a departure from good practice.  Its fix is four questions the tribunal asks at the first case management conference.

• Third, over instruction is real, but the expert is the person least able to see it.  The answer is not to remove counsel.  It is to record the assumptions and the final instructions the expert worked to, so the tribunal can see the boundaries; but in a departure from the Ciarb Report suggestions, we believe that the discussions that led to those instructions should stay private.

• Fourth, if only one recommendation is adopted, it should be the common data set.  The opposing experts agree, before they write their first reports, which documents and methods they will both use.  Over 80% of every group supports it.

• Fifth, the Ciarb Report asks things of experts too: read the existing guidance; make reports easy to navigate; do not seek private meetings with the tribunal; treat a change of mind as a strength.

The EWI would welcome hearing from arbitrators, judges, counsel and fellow experts who have tried any of the tools discussed in the Ciarb Report. That includes a common data set, a teach in, the four timetable questions or figures-as-figures directions. We are keen to know whether they worked or not. The Ciarb Report's authors have said they want the findings debated. The more that debate draws on real cases, the more useful the next edition will be. You can contact us at policy@ewi.org.uk.