International commercial arbitration is celebrated as the preferred method for resolving cross-border business disputes. Efficient, neutral, and enforceable in 172 countries under the New York Convention 1958: the argument for arbitration is compelling. But a growing body of critical scholarship asks a sharper question. When a corporation can bury the outcome of a dispute in a confidential process it helped design, is that justice, or is it just a better deal for the party with more power?
This post works through one of the central tensions in modern arbitration law: the relationship between confidentiality and accountability. The question is not whether arbitration is better than litigation in the abstract. It is whether the confidentiality architecture of the system creates structural cover for corporate misconduct, and what the legal framework should do about it.
What Confidentiality in Arbitration Actually Means
Confidentiality in arbitration is not a single rule. It is a patchwork of obligations that varies across institutions, governing laws, and individual contracts. The ICC Rules do not impose a general confidentiality obligation by default. The LCIA Rules do. The DIAC Rules include confidentiality provisions. The UNCITRAL Model Law, which forms the basis for UAE Federal Arbitration Law No. 6 of 2018, does not mandate confidentiality unless the parties agree to it.
What this means in practice is that the scope of confidentiality depends on what the contract says, which institutional rules apply, and which seat the parties chose. Tribunals sitting in England operate under a common law implied obligation recognised since Dolling-Baker v Merrett [1990]. The UAE's approach, particularly in DIFC-seated arbitration, reflects English law principles.
The Case for Confidentiality
The arguments for keeping arbitration private are real. Commercial parties have legitimate interests in keeping sensitive business information out of the public record. A dispute about a failed joint venture, a licensing breakdown, or a construction defect can expose pricing strategies and financial data that competitors would find valuable. Confidentiality also, in theory, encourages candour: parties may be more willing to engage openly in a private process than in one that ends up in a publicly accessible judgment.
The Problem: When Confidentiality Becomes Cover
The problem arises when confidentiality is not used to protect legitimate commercial sensitivity, but to prevent accountability for conduct that regulators, the public, or future contracting parties would want to know about.
The same feature that protects a company's pricing strategy can protect a company's fraudulent one. Confidentiality does not distinguish between secrets worth keeping and secrets worth exposing.
Consider the structure. A multinational corporation enters arbitration with a supplier who claims systematic non-payment and contract manipulation. The corporation, as a repeat player, has engaged its preferred institutional rules, retained experienced international counsel, and can absorb the cost of proceedings. The supplier cannot. Even if the supplier wins, the award is confidential. The corporation's conduct is not on public record. It can enter the next contract with the next supplier in exactly the same position.
The empirical literature on repeat players, developed by scholars including Marc Galanter and extended by researchers at the European University Institute, documents the systematic advantage that frequent arbitration users hold over one-time participants. Confidentiality amplifies this: repeat players accumulate knowledge of tribunal tendencies and procedural strategies that one-shot players cannot access.
The Access to Justice Dimension
Cost and confidentiality interact to compound the problem. ICC arbitration for a mid-complexity dispute costs on average between USD 500,000 and USD 1.5 million. For an SME whose claim falls near that range, the rational decision may be not to arbitrate at all, regardless of the legal merits. Confidentiality means that pattern of non-pursuit is invisible: there is no public record of the claims that were never brought.
This connects to a broader argument that arbitration does not merely resolve disputes neutrally, but shapes the distribution of legal outcomes in ways that track existing power asymmetries. The UNCITRAL Working Group III reform process has grappled with this question without yet producing a comprehensive answer.
What the Legal Framework Could Do
Three approaches are worth considering. First: mandatory disclosure for awards involving regulatory or public interest dimensions. Some jurisdictions already require this in investor-state arbitration. There is an argument for extending a version of this principle to commercial arbitration where the conduct at issue involves potential regulatory violations or patterns of contractual abuse. Second: third-party funding reform to correct the cost imbalance that drives smaller parties out of arbitration. Third: a database of anonymised awards, similar to what ICSID publishes for investment arbitration, which would reduce information asymmetry without exposing commercially sensitive detail.
A Genuinely Hard Question
There is no clean answer. Confidentiality serves real purposes, and mandatory disclosure would change the character of arbitration in ways that some legitimate commercial users would find unacceptable. The challenge is that the system as currently designed is most accessible to those who need accountability the least. Whether the legal framework will respond depends on whether the arbitration community is willing to treat structural fairness as a genuine design constraint, not an optional add-on.