When the curtain lifts on a mediation: Boghossian v IOP and the limits of “mediation privilege”

Most of the time, what happens in a mediation stays in the mediation. That promise of confidentiality is the oxygen of candid negotiation: parties can float ideas, make admissions, and explore compromise without fear that a judge will later pick over the conversation.
But every so often, litigation forces an awkward question: what exactly was agreed at the mediation (or what was said that induced someone to act)? When that question becomes a live issue, the court may have to look behind the curtain—not because confidentiality is unimportant, but because it cannot be used as a shield where justice requires the court to determine whether a binding agreement exists, what its terms were, or whether an exception applies.
A vivid example of this is Prof. Ardemis Boghossian v IOP Publishing Limited & Anor [2025] EWHC 3317 (IPEC) (judgment dated 17 December 2025).
The dispute that made the mediation “relevant”
The claim was about the alleged joint authorship (and copyright) of a scientific paper, and—critically—whether Prof Boghossian had consented to publication via an agreement reached in an EPFL-facilitated mediation in March/April 2016.
Here’s the key point for mediation confidentiality: both sides accepted that a mediation occurred and that an agreement was reached, but they disputed the agreement’s terms.
So the court had to decide (in the alternative, if joint authorship were proven): did the mediation agreement amount to consent to publish?
That is exactly the sort of issue that can pull the court into evidence about what happened around the mediation—even where parties would ordinarily insist it is “privileged”.
What evidence did the court look at?
Notably, the only written record of the mediation agreement was an email sent the next day (20 April 2016) by an EPFL representative (Ms Killias), setting out conditions and timing for publication.
Because the parties then argued over what was truly agreed, the court examined:
- the email record of the “deal” (including the 20 April email and later correspondence testing whether the “14 months” was a longstop or a review-point),
- witness evidence from the individuals involved (especially Prof Deveaud, whom the judge found credible and reliable),
- and the parties’ conduct after the mediation (including the absence of any timely objection when permission to publish was later granted).
Whatever label one applies—without prejudice, mediation confidentiality, mediation privilege—this is a practical instance of a court being willing (when necessary) to look at mediation-related material to resolve a dispute about the existence and meaning of the agreement said to have been reached.
The legal “how”: without prejudice protection and its exceptions
English law usually treats mediation communications as falling within the without prejudice rule, supported by contract confidentiality clauses and the general public policy of encouraging settlement.
But the rule has established exceptions. The classic starting point is Unilever v Procter & Gamble, which collects the main categories (including “unambiguous impropriety”).
Two exceptions matter most in practice when mediations later unravel:
1) The “what was the deal?” exception
Where it is necessary to determine whether a settlement was concluded and/or on what terms, the court can admit without prejudice material to decide that issue.
That principle is often discussed in the mediation context through Brown v Rice—a case frequently cited for the proposition that there is no magical, separate “mediation privilege” that blocks the court from doing its job when the existence/terms of a settlement are in issue.
In other words: confidentiality protects negotiation, but it cannot prevent the court from determining whether an enforceable agreement was reached (or what it meant) when that is the dispute before it.
2) The “something improper happened” exception
Without prejudice protection does not extend to communications that would act as a cloak for “unambiguous impropriety”—for example improper threats.
A modern illustration is Ferster v Ferster [2016] EWCA Civ 717, where the Court of Appeal upheld disclosure because the communication amounted to an unambiguously improper threat.
That is a higher threshold than mere sharp practice. But it matters because it underscores the point: without prejudice is not an absolute immunity.
A recent trend: “mediation privilege” arguments aren’t gaining much ground
There has been renewed debate in recent years about whether mediation deserves a bespoke, heightened privilege (beyond ordinary without prejudice principles). In Pentagon Food Group Ltd v B Cadman Ltd [2024] EWHC 2513 (Comm), the High Court considered these arguments and discussed the existing exception framework in the mediation context.
The practical message remains steady: courts prefer to work with the well-established without prejudice architecture, including its exceptions, rather than create an entirely new, sealed “mediation privilege” box.
What Boghossian v IOP teaches mediators and lawyers
This IPEC judgment is a reminder that mediation confidentiality is strong—but not invincible.
Three takeaways stand out:
- If the mediation outcome might matter later, write it down properly.
In Boghossian, the court was left with a thin written record and sharply divergent recollections—an invitation to satellite dispute. - Be precise with time limits and conditions.
“A lap time of max 14 months is to be foreseen” (in substance) became contested: longstop deadline or review milestone? Ambiguity is litigation fuel. - Remember that post-mediation conduct can speak loudly.
The court looked closely at what happened later—requests for permission, responses, and whether objections were made promptly.
The quiet irony
Mediation is designed to avoid adjudication. Yet when parties later litigate about the mediation itself, the court may have no choice but to examine communications that everyone assumed were safely confidential—because the dispute makes them necessary.
That is not a failure of mediation. It is a reminder that confidentiality is a tool for settlement, not a weapon for ambiguity.