Introduction
Clients confide in their lawyers on the assumption that those communications stay sealed. That assumption holds reasonably well within a single legal system. It collapses the moment a dispute crosses border. A communication that is fully protected in New York or Singapore can be wide open to disclosure in India, and the gap is rarely visible until a document request lands mid-arbitration.
This is the central problem of legal privilege in cross-border arbitration. Parties from different jurisdictions sit across the table, often before a tribunal seated in a third jurisdiction with its own privilege rules, each side carrying a different idea of what is protected. Before the tribunal can touch the merits of the dispute, it must answer a prior and frequently contested question: whose privilege law governs the documents in front of it?
This piece breaks down how legal privilege is defined in India, the UK, the US, and Singapore, how Indian law has shifted sharply on in-house counsel privilege, and how arbitral tribunals decide which jurisdiction’s privilege rules apply when multiple legal systems are in play.
What Legal Privilege Actually Covers
Legal privilege protects confidential lawyer-client communications from forced disclosure in legal proceedings. The principle sounds uniform. Its actual scope is not, and the differences matter most precisely when a dispute is multi-jurisdictional.
England: Legal Advice Privilege and Litigation Privilege
English law splits privilege into two distinct categories, as set out in Al Sadeq v. Dechert LLP, 2024 EWCA Civ 28. Legal advice privilege protects communications between a lawyer and client made for the sole or dominant purpose of giving or receiving legal advice. Litigation privilege protects communications created for existing or reasonably anticipated litigation, and this second category extends to third parties such as insurers and litigation funders, provided the dominant purpose of the communication is litigation-related advice.
Singapore: A Near-Mirror of English Law
Singapore’s privilege framework is codified in the Evidence Act, 1893, and it largely tracks the English position. Singapore recognises both legal advice privilege and litigation privilege, as affirmed in ARX v. Comptroller of Income Tax, 2016 SCC OnLine SGCA 1. Third-party communications receive protection under either category, as long as the dominant purpose test is satisfied, a position confirmed in Skandinaviska Enskilda Banken AB (Publ), Singapore Branch v. Asia Pacific Breweries (Singapore) Pte. Ltd., 2007 SCC OnLine SGCA 1.
United States: Broader Protection, No Dominant Purpose Test
US law uses two parallel doctrines: attorney-client privilege, which protects communications made for legal advice, and the work product doctrine, which shields material prepared in anticipation of litigation. The critical distinction from English and Singaporean law is that US courts do not impose a dominant purpose requirement. This makes the American standard meaningfully broader in practice.
Third-party extension under US law is layered. Work product protection extends to third parties without much friction. Attorney-client privilege is narrower and applies only where the third party is necessary for the lawyer to render advice, such as an accountant assisting counsel, a principle from United States v. Kovel (1961), or where parties share a common legal interest, such as co-defendants coordinating a joint defence, as recognised in Teleglobe Communications Corpn., In re (2007).
India: No Formal Split, and the Most Restrictive Standard
Indian law does not separate legal advice privilege from litigation privilege the way English and Singaporean law do. Instead, the Sakshya Adhiniyam, 2023 (which replaced the Indian Evidence Act, 1872) provides general statutory protection for confidential lawyer-client communications, with Indian courts applying a dominant purpose test broadly similar to England’s, as seen in Reliance Industries Ltd. v. SEBI, (2022) 10 SCC 181. Indian law does not extend privilege to third parties at all. Of the four jurisdictions compared here, India’s privilege regime is the narrowest in scope.
The In-House Counsel Question: India’s Position Has Just Been Settled
Whether privilege covers communications between an in-house lawyer and the company that employs them is where jurisdictions diverge most sharply, and it is also where Indian law has seen its most consequential recent development.
English law extends privilege to in-house counsel, but only when the lawyer is acting in a legal capacity and the communication concerns legal advice rather than purely business or administrative matters, as clarified in AAZ Applicant v. BBZ, 2016 SCC OnLine EWHC 25. The US follows a similar approach, grounded in Upjohn Co. v. United States, 449 US 383 (1981). Singapore codifies the same protection under Section 128-A of its Evidence Act.
India took a different and, until recently, uncertain path. That uncertainty was resolved by the Supreme Court in In Re: Summoning Advocates Who Give Legal Opinion Or Represent Parties During Investigation of Cases & Related Issues, 2025 INSC 1275, decided by a Bench of Chief Justice B.R. Gavai, Justice K. Vinod Chandran, and Justice N.V. Anjaria. The Court held that in-house counsel are not “Advocates” within the meaning of the Advocates Act, 1961, because they are full-time, fully salaried employees rather than independent legal practitioners. On that basis, the Court ruled that in-house counsel cannot claim the advocate-client privilege available under Section 132 of the Sakshya Adhiniyam, 2023 (BSA).
The Court’s reasoning centred on independence. It held that an in-house counsel, despite advising on questions of law, remains influenced by the commercial and business strategies of the employer and is institutionally obliged to protect the employer’s interests, which is incompatible with the independence that privilege is meant to safeguard. The judgment drew support from the European Court of Justice’s ruling in Akzo Nobel Ltd. v. European Commission, which similarly found that in-house lawyers lack the independence required for privilege under EU competition law, as well as Indian precedents including Rejanish K.V. v. K. Deepa and Bar Council of India v. A.K. Balaji.
This does not leave in-house communications entirely unprotected. The Court clarified that while Section 132 BSA privilege is unavailable, in-house counsel communications may still attract narrower protection under Section 134 of the BSA, which bars compelling disclosure of confidential communications made to a legal adviser before a court, subject to statutory exceptions such as the client offering themselves as a witness. This is a materially thinner shield than full advocate-client privilege, and it does not extend to communications from the employer to the in-house counsel, only to what the in-house counsel receives in that limited advisory role.
The result is a clear hierarchy across the four jurisdictions: the US offers the broadest protection with no dominant purpose test, generous third-party extensions, and full coverage for in-house counsel; the UK and Singapore sit in the middle with dominant-purpose-qualified in-house protection; and India now stands as the most restrictive jurisdiction, having expressly excluded in-house counsel from its core privilege protection.
For multinational companies with Indian subsidiaries, this creates a structural asymmetry. A US parent company’s in-house legal team can communicate under privilege. Its Indian subsidiary’s in-house counsel cannot claim the same protection for materially similar communications, even within the same corporate group and the same underlying dispute.
When Is Privilege Lost?
Even where privilege exists, it does not survive every circumstance. Across all four jurisdictions, the underlying logic is the same: privilege belongs to the client, so only the client’s conduct can extinguish it.
Under English law, privilege is lost where a party expressly waives it, shares the privileged material with an unprotected third party, or relies on a privileged document in the proceeding even partially, a position settled in Fulham Leisure Holdings Ltd. v. Nicholson Graham & Jones, (2006) 2 All ER 599. US law tracks the same three triggers, as reflected in Hearn v. Rhay (1975): express waiver, disclosure to an unprotected third party, or reliance on the substance of the privileged communication.
Indian and Singaporean law follow a similarly worded test. Privilege is lost when a party calls their own lawyer as a witness or relies on the privileged communication in the proceeding. The thread connecting all four systems is that privilege is forfeited only when the client, directly or indirectly, volunteers the protected material. Inadvertent disclosure, third-party leaks, or compelled production typically do not extinguish the privilege in the same way voluntary reliance does.
Whose Privilege Rules Apply in a Multi-Jurisdiction Arbitration
The comparative differences above become operationally significant the moment an arbitration involves parties, counsel, or a seat spanning more than one jurisdiction. The IBA Rules on the Taking of Evidence in International Arbitration, 2020 give tribunals the power to exclude privileged documents from evidence, but the Rules do not specify which jurisdiction’s privilege standard a tribunal should apply. That determination is left entirely to the tribunal’s discretion, decided case by case.
In practice, tribunals fall back on national laws because privilege is a creature of domestic legal systems with no internationally harmonised standard, an approach documented in Gary B. Born’s International Commercial Arbitration (3rd Edn., Kluwer Law International, 2020). The difficulty is that when parties come from different countries, applying “national law” immediately raises a conflict of laws problem: which nation’s law?
The Seat-of-Arbitration Approach
One method is to apply the privilege rules of the arbitral seat, on the reasoning that the seat is the legal anchor of the proceeding. This has the virtue of predictability, but it carries a real risk: if the seat’s law adopts a narrow privilege standard, parties from more protective jurisdictions may find communications they reasonably believed were shielded suddenly exposed. Where the seat itself is unclear or contested, this approach adds confusion rather than resolving it.
The Closest-Connection Approach
A second method looks to the law of the jurisdiction where the lawyer practises or where the client is based, an approach Born also discusses in International Commercial Arbitration. This tracks the parties’ actual expectations at the time the communication was made, which is a stronger justification than the seat-based approach. But it creates its own fairness problem in disputes involving multiple corporate entities across jurisdictions. Take the in-house counsel divide directly: a US company’s in-house legal advice is privileged, but its Indian subsidiary’s equivalent communications are not, following the 2025 INSC 1275 ruling. Applying different privilege standards to entities within the same arbitration, even within the same corporate family, raises genuine questions of equal treatment between the parties.
The Most-Favoured-Privilege Approach
A third school of thought argues tribunals should apply whichever standard is most protective among the connected jurisdictions, sometimes called the most-favoured-nation approach to privilege, discussed by Berger in “Evidentiary Privileges: Best Practice Standards versus/and Arbitral Discretion,” (2006) 22 Arb. Int. The fairness logic has surface appeal: why should one party be forced to disclose what the other can withhold? But privilege derives from national law, and granting a party protection it would not actually have under its own legal system, purely because the counterparty enjoys broader protection at home, is difficult to justify as a matter of principle. It also creates a real risk of exposing the resulting award to a public policy challenge at the enforcement stage.
There is no single formula that resolves this. Tribunals exercise broad discretion and frequently blend these approaches or construct a bespoke solution tailored to the specific parties and documents before them.
Confidentiality in Indian Arbitration: The Section 42A Framework
Privilege and confidentiality are related but distinct concepts, and India’s statutory confidentiality regime adds another layer relevant to cross-border disputes seated in India. Before 2019, the Arbitration and Conciliation Act, 1996 contained no express confidentiality provision for arbitral proceedings; confidentiality was statutorily recognised only for conciliation, under Section 75. Following the recommendations of the Justice B.N. Srikrishna-led High Level Committee in its 2017 report, the Arbitration and Conciliation (Amendment) Act, 2019 inserted Section 42A, which came into force on 30 August 2019.
Section 42A requires the arbitrator, the arbitral institution, and the parties to maintain confidentiality of all arbitral proceedings, with disclosure permitted only where necessary for the implementation and enforcement of the award. This is narrower than the High-Level Committee originally proposed, which had recommended broader exceptions, including disclosure necessary to protect a legal right. The narrower statutory exception has practical consequences: parties routinely place arbitral pleadings and records before courts in Section 9 interim relief applications, Section 34 challenges to awards, Section 37 appeals, and Section 11 applications for appointment of arbitrators, none of which Section 42A’s enforcement-only exception clearly covers. Section 42A also does not specify any penalty for breach, which limits its practical enforceability, and it does not extend to non-parties present during proceedings such as tribunal secretaries, transcribers, or experts.
For parties and counsel handling cross-border disputes with an Indian seat, this means confidentiality cannot be assumed to travel automatically into connected court proceedings, and privilege claims over the same documents may need to be asserted independently in each forum.
Conclusion
Legal privilege began as a simple, narrow protection against forced disclosure of lawyer-client communications. It has since had to absorb the complexity of multinational law firms, in-house legal departments, cross-border corporate structures, and international arbitration, none of which existed when the doctrine was first conceived. The result is a question that remains conceptually simple but has become operationally difficult: whose rules govern, and what counts as protected.
For India specifically, the position has moved in a clearly restrictive direction. The Supreme Court’s 2025 ruling on in-house counsel, read alongside the narrow confidentiality carve-out under Section 42A of the Arbitration and Conciliation Act, 1996, means Indian-connected parties in cross-border arbitration face a materially different privilege landscape than their counterparts in the US, UK, or Singapore. Until international arbitral practice converges on a clearer default rule, the safest course for parties is to address privilege expressly in the arbitration agreement rather than leave it to be litigated, document by document, once the dispute has already begun.
The principles examined in Sovereign Autonomy of Foreign-Seated Arbitration Law are closely connected to cross-border questions of legal privilege, where the applicable law of the seat can significantly influence confidentiality and disclosure obligations in international arbitral proceedings.
