Extending Arbitration Agreements to Third Parties under English Law

Principles, pitfalls, and practical drafting from TRW Law Firm

Snapshot

English law begins from a simple premise with far-reaching consequences: arbitration is consensual. The jurisdiction of an arbitral tribunal flows from the arbitration agreement, which is itself a contract. That starting point—privity and consent—makes English law cautious about dragging non-signatories into an arbitration. And yet modern commerce is multi-party and multi-contract: sponsors, SPVs, guarantors, assignees, trustees, funds, insurers, sub-contractors, and affiliates all swirl around the same transaction. When disputes arise, parties routinely ask: Can we bind (or be bound by) a third party to our arbitration clause?

This TRW analysis sets out the core English-law routes by which an arbitration agreement might extend beyond its signatories, the limits that courts and tribunals enforce, and the practical moves to bake clarity into your contracts and case strategy. We also flag the knock-on effects of the English Arbitration Act 2025 (notably the new default that the law of the seat governs the arbitration agreement) on non-signatory questions.

Bottom line: English law will extend arbitration agreements to third parties only on recognised legal bases (consent, statutory pathways, agency, assignment/novation, limited veil-piercing), and it will resist expansive doctrines that short-circuit consent. Careful drafting and disciplined pleading make the difference between an efficient single-forum resolution and a fractured, multi-jurisdictional fight.

1) First Principles: Consent, Privity, and the English Judicial Temperament

1.1 Consent is the cardinal rule

Arbitrations are “creatures of contract.” Tribunals take jurisdiction because the parties agreed. Non-signatories, by definition, did not sign. For English courts, that is the beginning and often the end of the analysis unless a recognised doctrine applies.

1.2 The cautionary approach

English courts have repeatedly warned against compelling strangers to arbitrate. They recognise the potential asymmetry (costs, forum, procedural law) and the due-process risk of binding a non-consenting party. Awards that rope in outsiders are vulnerable to jurisdiction challenges and enforcement headwinds.

1.3 Dallah as a north star (and a trans-Channel contrast)

The well-known saga involving Dallah and Pakistan remains instructive. French courts took a holistic view of conduct and relationships to bind a non-signatory; English courts took a strict consent view and refused to enforce. The lesson: seat and enforcement geography matter. If you seat in London—or plan to enforce in England—you must make your non-signatory theory fit within English principles.

2) The English-Law Toolset: When Extension Can Happen

English law does recognise a number of orthodox routes by which an arbitration agreement reaches a third party. Each requires fact-specific discipline.

2.1 Express joinder and consolidation (by agreement)

• Joinder: Many institutional rules (e.g., LCIA, ICC) allow joinder if the third party consents (and often if all existing parties do). Without consent, tribunals cannot compel a stranger to join.
• Consolidation: Section 35 of the 1996 Act permits consolidation only with party agreement or under rules/clauses that authorise it. Consolidation may practically align multiple cases but still rests on consent.

Practice tip: If multi-party risk is foreseeable, write joinder/consolidation mechanics into the contract suite at the outset—don’t rely on later goodwill.

2.2 Contracts (Rights of Third Parties) Act 1999 (the “1999 Act”)

• Substantive right + arbitrate condition: A third party the contract identifies (by name, class, or description) may enforce a benefit under Section 1(1). If the contract makes that enforcement subject to arbitration, Section 8(1) treats the third party as a party to the arbitration agreement for disputes about the enforcement of that benefit.
• Scope is narrow: The third party is not magically a signatory for all purposes. It can be required to arbitrate to pursue its 1999 Act benefit, but it cannot necessarily force signatories into other arbitrations or run broader anti-suit strategies unless the contract expressly gives that power (see also Section 8(2)).

Practice tip: If you intend true symmetry—i.e., the third party both must and may arbitrate on all disputes tied to its benefit—state this expressly. Otherwise you may create a one-way valve.

2.3 Trust beneficiaries and arbitrability

Historically, trust disputes were considered the courts’ domain. More recent High Court remarks indicate no general prohibition on arbitrating trust disputes. This opens the door (cautiously) to extending arbitration agreements to beneficiaries where instruments and identification requirements are satisfied.

Practice tip: In trust-heavy structures, draft a trust arbitration clause that: (i) identifies beneficiary classes; (ii) anchors disputes to arbitration; and (iii) sets out a representation mechanism for unborn/minor beneficiaries consistent with due process.

2.4 Agency

If an agent signs a contract on behalf of a principal, the principal is bound—including by the arbitration clause. This is not an “extension”; it is orthodox attribution of consent. The fight is factual: did the agent act with authority (actual, ostensible, or ratified)?

Practice tip: Use authority certificates and no-oral-modification clauses. They reduce later “no consent” defences.

2.5 Assignment

An assignee takes the benefit of a contract and—under English law’s separability logic—generally takes the arbitration clause with it (“the burden that travels with the benefit” for dispute resolution). Anti-assignment clauses can derail this, including statutory assignments if the drafting is sufficiently clear.

Practice tip:

• In the assignor–assignee deed, include a deeming provision that the assignee assumes the arbitration agreement and appoints the same seat/rules.
• In the head contract, pair any anti-assignment clause with a permitted assignment carve-out that conditions assignment on the assignee’s express assumption of the arbitration agreement.

2.6 Novation

Novation replaces the old contract with a new one, transferring both benefits and burdens—including the arbitration clause—with consent of all parties. Failure to comply with no-oral-modification or other formalities can render a supposed novation void.

Practice tip: Use a short-form tripartite novation that restates the arbitration clause in full to avoid arguments about survival.

2.7 Statutory pathways

• Death and personal representatives: The 1996 Act provides that, absent contrary agreement, an arbitration agreement survives death and is enforceable by/against the deceased’s personal representatives.
• Insolvency: Where a trustee in bankruptcy adopts a contract, the arbitration clause may become enforceable by or against the trustee for connected matters—subject to the contract’s anti-assignment architecture.

Practice tip: Audit insolvency and succession exposure in your counterparties and put notification and step-in mechanics around the arbitration clause.

2.8 Piercing the corporate veil (rare, but real)

In exceptional cases—typically to prevent evasion of existing obligations—courts may treat the controller and the controlled entity as one, binding the “alter ego” to the arbitration clause. Courts stress that ordinary corporate-group relationships are not enough; there must be impropriety directed at frustrating the law.

Practice tip: Keep this as a last resort theory, and build a tight evidentiary record focusing on evasion rather than mere ownership.

3) What English Law Doesn’t Do: The Rejections and the Red Lines

3.1 The “group of companies” doctrine

Unlike some civil-law jurisdictions, English law does not adopt a free-standing “group of companies” doctrine under which affiliates are bound by implication. Mere participation in negotiations, performance, or termination by a group member does not, without more, create consent.

Consequence: To bind an affiliate, you need a contractual hook (e.g., that affiliate is a party, guarantor, assign/assume, or an express “affiliates bound” clause), or one of the orthodox doctrines above.

3.2 “Holistic” extension absent a doctrinal path

The French-style holistic approach (looking across conduct to infer consent) is not the English approach. English courts will ask: What doctrine are you invoking? Agency? Assignment? Estoppel? Veil-piercing? 1999 Act? If not, extension will likely fail.

3.3 Awards binding strangers

Tribunals that stride ahead and issue awards against non-signatories absent a recognised basis risk annulment at the seat and non-recognition at enforcement. Strategically, that can squander years.

4) The 2025 Act’s Quiet but Important Relevance

The English Arbitration Act 2025 amends the 1996 Act and, among other things, inserts a default rule (new Section 6A): the law of the seat governs the arbitration agreement unless the parties expressly provide otherwise. Why does that matter for non-signatories?

• Choice-of-law fights shrink: Many non-signatory theories (e.g., assignment scope, separability, non-signatory estoppel) depend on which law governs the arbitration agreement. The new default reduces satellite litigation about that law in London-seated contracts.
• Drafting precision still wins: If you want New York law (or another law) to govern the arbitration clause in a London-seated contract—for example, to access US-style estoppel theories—you must say so expressly. Conversely, if you want the English default, maintain silence—or say it expressly for belt and braces.

TRW view: For most cross-border deals in our Dhaka–Dubai–London corridor, the seat-law default is welcome; it predicts how English courts will treat non-signatory issues. If you intend to rely on foreign-law doctrines expanding non-signatory reach, make an explicit clause-law choice.

5) Drafting Playbook: Bake in or Fence Out Non-Signatory Reach (Pick One)

The best time to win a non-signatory fight is before it starts—in the clause bank. Here are battle-tested options you can copy into your templates.

5.1 If you want to extend reach (multi-party, multi-contract ecosystems)

1. Affiliates Bound Clause

“Each Party enters into this Agreement on its own behalf and, to the extent permitted by law, as agent for its Affiliates. Each Party procures that its Affiliates comply with and are bound by the Arbitration Agreement in Clause [X] in respect of any dispute arising out of or in connection with this Agreement to which such Affiliate is a party in interest.”

Pair with a definition of Affiliate and a notice mechanism for when an Affiliate’s acts give rise to a dispute.

2. Third-Party Beneficiary with Symmetric Arbitration (1999 Act + express symmetry)

“Any Person expressly identified as a Third-Party Beneficiary may enforce the benefit conferred on it only by arbitration under Clause [X]. Each Party agrees that such Third-Party Beneficiary may compel arbitration and may be compelled to arbitrate any dispute relating to that benefit.”

3. Assignment/Novation Assumption

“No assignment is effective unless the assignee executes a deed assuming all obligations and is deemed a Party to the Arbitration Agreement (seat, rules, and governing law of the arbitration agreement as stated).”

4. Joinder & Consolidation Mechanics

• Incorporate rules that enable joinder with streamlined consent (LCIA/ICC language) and pre-agree to consolidation across related agreements sharing the same seat/rules.

4. Guarantee/Direct Agreement

• Put the guarantor or parent in privity with a short, standalone arbitration clause mirroring the main contract (seat, rules, clause-law).

4. Trust Instruments

• Expressly submit beneficiary disputes to arbitration, appoint a representative procedure consistent with due process, and identify classes of beneficiaries.

5.2 If you want to fence out non-signatory reach (clean bilateral arbitration)

1. No Third-Party Rights

“A person who is not a Party has no right to enforce any term of this Agreement. Section 1(1) of the Contracts (Rights of Third Parties) Act 1999 is excluded.”

2. No Affiliates Bound

“Affiliates of a Party shall not be deemed parties to the Arbitration Agreement absent a written accession agreement executed by such Affiliate.”

3. Strict Anti-Assignment

“No assignment of any right is effective without prior written consent, and any purported assignment shall be void. This restriction applies to assignments by operation of law to the maximum extent permitted.”

4. Clause-Law Express Choice

• State that the arbitration agreement is governed by English law (if you want to avoid imported non-signatory doctrines) or pick another law expressly if you want them.

TRW drafting note: Think in families of contracts. Multi-contract projects often fall over because documents in the “suite” pick different seats/rules. Harmonise them or map an interop clause that allows consolidation across the suite.

6) Litigation & Arbitration Strategy: Making (or Beating) the Extension Case

6.1 If you are seeking to bind a non-signatory

• Pick a doctrine and prove its elements: agency (authority trail), assignment (effective transfer + anti-assignment carve-outs), novation (formalities), 1999 Act (identification + benefit + Section 8 wording), veil-piercing (evasion evidence).
• Build contemporaneous evidence: board minutes, emails showing assumption, undertakings, guarantee language, performance conduct compatible with party status.
• Seat and clause-law: leverage the 2025 Act default where English-law treatment favours your theory; or draft explicit clause-law if you need a foreign-law doctrine.
• Procedural route: consider joinder under institutional rules if the third party will consent. If not, frame a jurisdiction submission to the tribunal laying out your doctrinal basis and be prepared for a Section 67 challenge risk if you overreach.

6.2 If you are resisting extension

• Hammer consent: no signature, no authority, no novation formalities, anti-assignment engaged, 1999 Act excluded or limited in scope.
• Separate personhood: push back against group-of-companies reasoning; remind the tribunal of English law’s rejection of that doctrine.
• Seat leverage: in London, emphasise strictness of English control on non-signatory reach and the annulment risks of adventurous awards.
• Parallel court moves: consider a stay or anti-suit in support of the arbitration you did agree to—but beware asymmetries with partial 1999 Act rights.

7) Sector-Specific Notes

7.1 Construction & Infrastructure (EPC, O&M, supply chains)

• Use direct agreements and step-in clauses to bring lenders/employers/contractors into privity with a mirror arbitration clause.
• Pre-wire joinder/consolidation across EPC, supply, and services contracts—same seat/rules to avoid incompatibility.

7.2 Energy & Resources (PSCs, offtake, JVs)

• Combine affiliate guarantees with clean arbitration clauses; avoid relying on “group involvement” to imply consent.
• For JVs with layered agreements, adopt a master dispute resolution clause that permits consolidation.

7.3 Finance & Derivatives

• Assignment is routine. Pair permitted transfer language with assumption of the arbitration clause and an express clause-law selection to avoid surprises.

7.4 Private Wealth & Trusts

• If arbitrability of trust disputes is intended, say so and create a procedural representation mechanism for beneficiaries (including those unborn) compatible with fairness.

7.5 Technology & Data

• Identify the cloud/host/custodian reality. If third-party providers are key to performance, give yourself joinder hooks (with consent parameters) and Section 44 court-support pathways in the seat to reach non-parties for evidence preservation.

8) Common Traps (and How to Avoid Them)

• Mismatched clause families: Different seats/rules in related contracts cripple consolidation and joinder. Harmonise.
• Vague third-party language: “Affiliates may benefit” without saying how and where they arbitrate yields fights. Be explicit.
• Silent assignments: Transfers without an assumption deed invite “we didn’t consent” defences.
• Overreliance on conduct: English law will not infer consent just because an affiliate was “in the room”.
• Award overreach: Tribunals that bind strangers without a recognised basis create set-aside risk. Calibrate.

9) The Enforcement Lens

If your strategy depends on enforcing in England and Wales, expect strict scrutiny of non-signatory theories. If you plan to enforce elsewhere, map the comparative doctrine: some jurisdictions adopt expansive non-signatory doctrines (group of companies; implied consent from conduct). Your seat choice, clause-law, and asset map should be coordinated.

TRW note: Where enforcement will likely occur in multiple regions, consider a belt-and-braces approach—seat in London for procedural robustness; add express clause-law if you need a doctrine recognised abroad; and put direct agreements/guarantees in place to create privity with target assets.

10) Board-Ready Checklist (Print & Use)

• Have we harmonised seat/rules across the contract suite?
• Do we state expressly the law governing the arbitration agreement (post-2025 Act)?
• If third-party participation is likely, do we have joinder and consolidation mechanics?
• Are assignees forced to assume the arbitration clause via deed?
• Do guarantees and direct agreements carry mirror arbitration clauses?
• Have we excluded or tailored the 1999 Act as desired?
• For trusts/beneficiaries, do we have an arbitration and representation mechanism?
• Are no-oral-modification and anti-assignment clauses tight (and deliberately carved out where needed)?
• Do we have a joinder consent plan and a litigation map if consent is refused?
• If we must argue extension, which doctrine fits—and do we have the documents to prove it?

11) TRW’s Practical Packages

We turn the above into actionable documents and strategies for clients operating across Bangladesh, the GCC, and the UK:

• Clause-bank refresh for multi-party deals (affiliates, assignment, joinder, consolidation, trust/beneficiary language).
• Suite harmonisation for major projects (one seat/rules; interop consolidation clauses).
• Assumption and novation deeds (short-form tripartite templates).
• Non-signatory litigation memos tailored to London-seated cases (agency, assignment, veil-piercing).
• Enforcement mapping (where you’ll actually collect; align seat, clause-law, and doctrine exposure).

For an overview of our cross-border arbitration capability, visit our page on International Arbitration & Enforcement.

12) Conclusion

English law’s fidelity to consent and privity is not hostility to commercial reality; it is a demand for clarity. If you want third parties bound, say so in the paperwork and route through recognised doctrines. If you want a clean bilateral arbitration, fence out third-party rights and transfers that bypass consent. Either way, the worst outcome is ambiguity.

As disputes become more networked—across affiliates, trustees, funds, and platforms—the parties who win are those who engineer the forum in advance. That means aligned seats and rules, express clause-law choices, joinder/consolidation pathways, and assignment/novation deeds that leave nothing to “implication”.

TRW is here to help you convert principle into practice—from Dhaka to Dubai to London.

Contact TRW Law Firm

Phones (Bangladesh)
+8801708000660 · +8801847220062 · +8801708080817

Emails
[email protected] · [email protected] · [email protected]

Offices

• Dhaka: House 410, Road 29, Mohakhali DOHS
• Dubai: Rolex Building, L-12 Sheikh Zayed Road
• London: 330 High Holborn, London WC1V 7QH, United Kingdom

This publication is for general guidance only and does not constitute legal advice. For advice on specific transactions or disputes, please contact TRW’s international arbitration team.