Come Together? Acer v Nokia and the Contractual Turn of FRAND Arbitration

"The readiness is all." - Hamlet

Sometimes a judgment does not change the melody of FRAND law. It changes the instrument through which that melody may be played. Not every player reads the same score, and not every performance is equally faithful to the underlying composition. What the UK Court of Appeal has done in Acer v Nokia is reassign one voice in the ensemble - without rewriting the work.

The Court of Appeal's decision in Acer Inc. v. Nokia Technologies Oy is one of those judgments. It is not merely another episode in the now familiar English line of global FRAND/RAND cases. It is, more precisely, a decision about the legal form through which the FRAND/RAND undertaking may be performed: court determination, yes; but also, in the right circumstances, arbitration.

The point matters. The dispute concerned Nokia's video codec SEP portfolio, declared essential to ITU-T H.264/AVC and H.265/HEVC standards. Because the undertaking arose under ITU-T, the case was technically about RAND rather than FRAND, although nothing turned on that distinction for the purposes of the Court's analysis. Throughout this post, FRAND is used as the primary shorthand; RAND is retained where the ITU-T context demands precision. Acer and ASUS accepted that they needed global licences. Nokia had launched infringement proceedings seeking exclusionary relief in several jurisdictions, including Germany, the UPC, the United States, Brazil and India. Acer and ASUS then brought implementer-led proceedings in England seeking final RAND determinations and interim licence declarations. The High Court accepted jurisdiction, refused Nokia's stay application, and granted interim relief. The Court of Appeal upheld jurisdiction but allowed Nokia's appeal on the stay. The key reason was Nokia's "Adjustable Licence Offer": an immediate interim licence, adjustable once an arbitral tribunal determined the final RAND terms.1

That is the surface of the case. Its deeper importance lies elsewhere. The Court of Appeal accepted that an offer of an immediate licence, with final terms to be determined by an ICC arbitral tribunal, could itself be an objectively RAND offer capable of acceptance. Refusal of that offer could therefore deprive the implementer of the ability to invoke the English court's declaratory jurisdiction for a competing RAND determination.2

Not compulsory arbitration, but contractual performance

The natural objection is simple: arbitration is consensual. The ITU-T Declaration did not contain an arbitration clause. Acer and ASUS did not agree to arbitration. Therefore, one might say, Nokia could not transform its RAND obligation into an arbitral obligation by unilateral drafting.

That was essentially the High Court's view. The Court of Appeal, however, took a more subtle route. It did not say that the ITU-T Declaration contains an arbitration clause. Nor did it say that national courts may rewrite SDO patent policies in order to solve structural problems in global SEP litigation. On the contrary, Arnold LJ expressly accepted that courts should approach that issue with caution.3

The decisive point was different. Acer and ASUS were not compelled to arbitrate. They were offered a choice: accept an immediate licence whose final terms would be set by an independent arbitral tribunal, or refuse that licence and bear the legal consequences of refusing what the Court treated as a RAND offer. That is not arbitration imposed by fiat. It is what this post characterises - as a matter of doctrinal construction, not explicit holding - as arbitration embedded in the performance of a contractual undertaking. The Court of Appeal did not use precisely that language. It reasoned from the specific features of Nokia's offer to the specific outcome of the stay.4

That said, the choice was formally free without being commercially neutral. Nokia's infringement proceedings remained live in multiple jurisdictions, with the attendant risk of exclusionary relief in Germany, the UPC, the United States and elsewhere. To refuse the offer was to refuse licensed access while that exposure continued. A technically uncoerced refusal is not the same as a genuinely unconstrained one.

The contractual foundation of the reasoning should not be missed. The judgment records that, on the findings of Swiss law, the ITU-T Declarations gave rise to a contract for the benefit of implementers; it then treats the established ETSI FRAND jurisprudence as applicable in substance.5 That is not a marginal detail. It explains why the inquiry is not merely whether a national court should manage parallel proceedings, but whether the SEP holder has offered a legally adequate mode of performance of an undertaking owed, directly or indirectly, for the benefit of implementers.

This is where the decision becomes doctrinally interesting. FRAND is too often discussed as if it were a remedial algorithm: offer, counteroffer, security, injunction, rate. But the more fundamental question is contractual. What has the SEP holder promised? To whom? Through what legal mechanism may that promise be performed?

If the FRAND/RAND undertaking is a contractual obligation to make access available on objective terms, then neither party owns the process entirely. The implementer has a right to access on FRAND/RAND terms, not necessarily a right to dictate the adjudicatory forum whenever a credible FRAND/RAND mechanism has been offered. Equally, the SEP holder has not preserved an ordinary right to exclude without qualification: it must make available a genuine route to licensed access.

On that view, arbitration does not displace the FRAND obligation. It may concretise it.

The judgment should not be overread

The decision is pro-arbitration, but it is not a blank cheque.

First, Nokia offered an immediate global interim licence. The Court expressly left open whether a mere offer to arbitrate, without an immediate licence, would be sufficient. That reservation is essential. Arbitration cannot be used as a waiting room while exclusionary proceedings continue to generate commercial pressure. If arbitration is to serve FRAND, it must preserve access while the rate is being determined. The judgment does not resolve what happens to pending infringement actions in other jurisdictions while the arbitral process runs; that question is left open and will be among the first issues tested in subsequent cases.6

Second, the Court's analysis depended on the absence of any substantiated objection to the proposed ICC arbitration. Nokia proposed ICC Rules, a neutral seat and place not in either party's home jurisdiction, three arbitrators, disclosure of comparable licences, and, shortly before the appeal, a public non-confidential summary of the award's main points. Acer and ASUS did not make a counterproposal for a different form of arbitration. The Court expressly left open what would happen if an implementer had a legitimate and substantiated objection to the patentee's proposed arbitral mechanism.7 It gave no full guidance on what such an objection might look like. Possible candidates include a proposed seat in a jurisdiction with an uncertain enforcement record, institutional rules offering no workable mechanism for disclosure of comparable licences, a confidentiality regime so absolute as to preclude any published reasoning, or a tribunal-constitution process that structurally advantages one side. These are not abstract concerns. They are likely to be the next battleground.

Third, transparency remains unfinished business. Arnold LJ rightly noted that ICC arbitration does not automatically impose confidentiality in the way some assume, and added that there is no good reason why an arbitral tribunal determining RAND terms should not publish its award subject to necessary redactions. That is a crucial point. FRAND arbitration will not gain legitimacy if it is perceived as a black box in which global licensing norms are privately manufactured and then invoked publicly without scrutiny.8

Fourth, asymmetry remains. The Court was candid about it. A SEP holder can bring infringement proceedings; an implementer, absent such a claim, typically seeks declaratory relief. The patentee may also choose among objectively RAND terms the set most favourable to itself. That asymmetry is real. But the answer is not to deny arbitration as such. The answer is to police the conditions under which arbitration is proposed.9

From paranoia to procedure

Last year, I suggested on this blog that arbitration could offer a partial answer to the "paranoid" landscape of SEP litigation: parallel proceedings, anti-suit tactics, inconsistent rulings and strategic escalation. But the caveat was, and remains, decisive: arbitration must be properly structured.10

Acer v Nokia gives that intuition judicial architecture. It does not say that arbitration is always preferable to court determination. It says that an arbitral route may be a legitimate mode of performance where it gives the implementer immediate access, independent determination, procedural fairness, and enforceability.

The decision should be read alongside the established English line of global FRAND cases, of which Unwired Planet v Huawei, Optis v Apple and the recent interim-licence decisions are the most significant antecedents. Those cases established the English court's willingness to determine global FRAND rates and to organise interim access pending determination. Acer v Nokia does not displace that jurisdiction. It introduces a further variable: where the patentee itself proposes a credible arbitral route to licensed access, the court may stay its own proceedings rather than assert its rate-setting authority. The relationship between that variable and the existing doctrine will take time to settle.11

The Mannheim signal: PMAC as procedural gravity

The point is not purely English. Almost at the same moment, a parallel signal came from Mannheim in the global dispute between ZTE and Samsung. In the English leg of that dispute, Meade J had determined a FRAND lump-sum balancing payment of $392 million for a global cross-licence and observed that the proliferation of litigation was symptomatic of a dysfunctional system whose critical problem was the absence of a dispute-resolution mechanism within the ETSI rules.12

The UPC Mannheim Local Division then did something revealing. Instead of simply adding another adjudicative layer to the conflict, it issued a settlement proposal suggesting two possible global cross-licence outcomes: $640 million for a licence ending on 31 December 2028, or $730 million for a licence ending on 31 December 2029. Failing agreement, the panel suggested mediation under the auspices of the PMAC, with a stay of the pending UPC cases upon the parties' mutual request and a mechanism for the UPC to confirm any resulting settlement. The parties were ordered to comment on the proposal and the proposed PMAC mediation by 31 May 2026.13

The parallel with Acer v Nokia is obvious, but it should not be overstated. Acer concerned arbitration embedded in an immediate licence offer and used to justify a stay of implementer-led English RAND proceedings. The Mannheim order is different: it is not a binding FRAND determination, not an arbitral award, and not a compulsory referral to mediation. It is a court-assisted invitation to convert a litigation spiral into a negotiated solution. Yet the institutional intuition is the same. When the parties agree that a licence must exist but fight globally over price, courts increasingly see procedure not merely as case management, but as a means of restoring the contractual economy of FRAND.

From a contract-law perspective, this is not an accident. FRAND is an obligation whose content is binding but not fully self-executing. It therefore requires mechanisms of concretisation: negotiation first, then judicial determination, arbitral determination, expert determination or mediation where the parties can be brought to a workable procedural architecture. The PMAC rules make that architecture visible. The Mediation Rules expressly contemplate FRAND disputes and list issues such as portfolio sampling, cross-licensing, temporary licensing terms, royalty base and range, and royalty methodology. The Arbitration Rules contain a dedicated section on FRAND disputes and even address parallel proceedings, including stays or undertakings not to seek or enforce injunctions where the parties so agree.14

Where arbitral institutions matter

The ICC was the institution proposed in Acer v Nokia, and nothing in the judgment suggests that this was inappropriate. But ICC arbitration is not the only credible route. WIPO has developed dedicated ADR tools for SEP/FRAND disputes, including model submission agreements for mediation, arbitration, expedited arbitration and expert determination; it also reports having administered a significant number of SEP/FRAND-related disputes.15

The UPC Patent Mediation and Arbitration Centre is also becoming strategically important. PMAC is no longer a merely prospective institution: its rules and legal documents now include Mediation Rules and Arbitration Rules expressly addressing FRAND/RAND disputes. This matters because specialised ADR is most useful where it does not merely provide a room for bargaining, but supplies a procedural grammar: confidentiality rings, sampling, royalty methodology, temporary terms, parallel proceedings, and a route from technical evidence to a reasoned outcome.16

PMAC may be particularly interesting in the European context. The UPC itself is not yet the natural home for global FRAND rate-setting. Its jurisdiction, remedies and institutional role remain contested. But a specialised arbitral centre connected to the UPC ecosystem could become a useful forum precisely because it is not a court trying to stretch territorial patent adjudication into a global licensing machine. It could offer something more modest and potentially more effective: a neutral mechanism for contractual determination.

WIPO offers experience and international neutrality. ICC offers institutional credibility and procedural robustness. PMAC offers technical specialisation and proximity to the European patent litigation environment. Other institutions may also be suitable. The institution is less important than the architecture: neutrality, expertise, disclosure, procedural economy, confidentiality only where justified, enforceability, and a publishable reasoned outcome.17

A decision for both sides

SEP holders should not read Acer v Nokia as permission to weaponise arbitration. A patentee who wishes to rely on an arbitral mechanism should make the offer look like what it claims to be: a route to FRAND access. That means an immediate licence or standstill protecting market access; a neutral and reputable institution; a balanced tribunal-selection process; meaningful disclosure of comparables; a mechanism for dealing with confidentiality; continuity of evidence and pleadings where litigation has already begun; and publication, at least in redacted form, of the reasoning on valuation and principle.

Implementers should not read "consent" as a magic word either. If an arbitration proposal is credible, refusal without engagement may increasingly look like hold-out. The better response is not ritual rejection, but procedural counter-design: propose WIPO, PMAC, ICC, LCIA or another credible forum; identify the seat; define disclosure; ask for a reasoned award; protect confidential licence data without suppressing legal reasoning; and insist that the arbitral process preserve access during the determination.

That is the practical lesson of the case. FRAND is not served by allowing patentees to choose an opaque forum that maximises pressure. Nor is it served by allowing implementers to insist on court determination as a tactical refuge when a genuine arbitral route to licensed access is available. Both forms of opportunism are contrary to the contractual economy of FRAND.

The contractual turn

The Court of Appeal's decision fits within a broader contractual understanding of FRAND/RAND. The undertaking is not a free-floating regulatory slogan. It is not merely a competition-law defence. Nor is it an automatic licence formed by implementation of a standard. It is a commitment to make access available on objective terms, enforceable by those for whose benefit the undertaking was made.

From a civil-law perspective, the point may be expressed with some precision. Under the ETSI framework governed by French law, the FRAND undertaking is naturally read as a stipulation pour autrui: the standard-setting organisation receives the undertaking, the SEP holder promises to make FRAND licences available, and future implementers constitute an open but determinable class of beneficiaries. The beneficiary does not thereby become a party to an already completed licence; it acquires a direct right to performance of the undertaking, whose precise terms may require negotiation, judicial determination, or arbitral determination. That distinction - between obligation and downstream licence - is exactly what allows arbitration to be conceived as a mode of performance rather than as a substitute source of obligation.18

Once that is accepted, the determination of FRAND/RAND terms is not necessarily the monopoly of a national court. A court may determine those terms. An arbitral tribunal may determine them. The essential question is whether the chosen mechanism faithfully performs the underlying obligation.

That is why Acer v Nokia is important. It shifts the debate from forum prestige to contractual performance. The English court remains central, but not exclusive. Arbitration becomes not an escape from law, but one possible juridical form through which the FRAND undertaking is given effect. The turn is real. But it has been made on facts unlikely to recur without deliberate structuring: an immediate global interim licence, a credible institutional proposal, the absence of any substantiated counter-proposal, and a commitment to publish at least a non-confidential summary of the outcome. Remove any of those elements, and the result may be different.

This is not the end of the story. The Supreme Court may yet be asked to intervene. Future cases will have to decide what counts as a legitimate objection to arbitral determination, whether an implementer can itself rely on an arbitral offer against an unwilling SEP holder, and how transparent arbitral awards must be in order to sustain confidence in FRAND outcomes.

For now, the message is clear enough. Arbitration is no longer merely a polite academic suggestion for SEP disputes. Properly structured, it may be part of the law of FRAND performance.

Not a panacea. Not a weapon. A procedure.

• 1Acer Inc. v. Nokia Techs. Oy, [2026] EWCA Civ 564, [1]-[17], [91]-[94] (Eng. C.A.).
• 2Id. [84]-[85], [91].
• 3Id. [81]-[83].
• 4Id. [67]-[85].
• 5Id. [21]-[24]. The Court recorded that, on the High Court's findings of Swiss law, the ITU-T Declarations gave rise to a contract for the benefit of implementers and that the ETSI FRAND jurisprudence was applicable by analogy.
• 6Id. [74]-[77], [85].
• 7Id. [29], [79]-[80].
• 8Id. [79]; ICC Rules of Arbitration arts. 22(3), 32(2), 35 (2021), https://iccwbo.org/dispute-resolution/dispute-resolution-services/arbitration/rules-procedure/2021-arbitration-rules/. Article 22(3) permits the tribunal, upon request, to make confidentiality orders; Article 32(2) requires the award to state reasons; Article 35 governs notification and deposit of awards.
• 9Acer, [2026] EWCA Civ 564, [86]-[88].
• 10Matthieu Dhenne, SEP, FRAND, and Arbitration: A “Paranoid” Quest for Harmony in Global Patent Disputes, Kluwer Patent Blog (Apr. 29, 2025), https://legalblogs.wolterskluwer.com/patent-blog/sep-frand-and-arbitration-a-paranoid-quest-for-harmony-in-global-patent-disputes/.
• 11Unwired Planet Int'l Ltd. v. Huawei Techs. (UK) Co., [2020] UKSC 37, [58]-[64], [88]-[90] (U.K.); Optis Cellular Tech., LLC v. Apple Retail UK Ltd., [2025] EWCA Civ 552, [20]-[21], [143] (Eng. C.A.); Lenovo Grp. Ltd. v. Telefonaktiebolaget LM Ericsson, [2025] EWCA Civ 182, [91], [115], [129], [136], [156]-[157] (Eng. C.A.).
• 12Samsung Elecs. Co. v. ZTE Corp., [2026] EWHC 999 (Pat), [1]-[7], [48], [576] (Eng.).
• 13Christina Schulze, The UPC Wants to Bring ZTE and Samsung to the Negotiating Table, JUVE Patent (May 13, 2026), https://www.juve-patent.com/people-and-business/the-upc-wants-to-bring-zte-and-samsung-to-the-negotiating-table/; Samsung Elecs. Co. v. ZTE Deutschland GmbH, UPC_CFI_189/2025, Order (Settlement Proposal) at 2-4 (UPC CFI, Mannheim Loc. Div. May 13, 2026), https://ipfray.com/wp-content/uploads/2026/05/ZTE_SAMSUNG_SETTLEMENT_Proposal_signed.pdf.
• 14Patent Mediation & Arbitration Ctr. of the Unified Patent Court, Mediation Rules arts. 5(1), 26 (adopted Apr. 24, 2026), https://www.pmac-upc.org/sites/default/files/upc_documents/D_AC_06_04112025_rev.2%20Mediation%20Rules_EN.pdf; Patent Mediation & Arbitration Ctr. of the Unified Patent Court, Arbitration Rules arts. 48-50 (adopted Apr. 24, 2026), https://www.pmac-upc.org/sites/default/files/upc_documents/D_AC_07_04112025_rev.2%20Arbitration%20Rules_EN.pdf.
• 15WIPO Arbitration & Mediation Ctr., WIPO ADR for SEP/FRAND Disputes, WIPO, https://www.wipo.int/amc/en/center/specific-sectors/ict/frand/ (last visited May 13, 2026). WIPO states that it has administered some 95 SEP/FRAND-related disputes and provides model submission agreements for mediation, arbitration, expedited arbitration, expert determination, and multi-tier procedures.
• 16Patent Mediation & Arbitration Ctr. of the Unified Patent Court, Rules & Legal Documents, PMAC, https://www.pmac-upc.org/en/rules-legal-documents (last visited May 13, 2026); Patent Mediation & Arbitration Ctr. of the Unified Patent Court, Arbitration Rules arts. 1, 48-50 (adopted Apr. 24, 2026), https://www.pmac-upc.org/sites/default/files/upc_documents/D_AC_07_04112025_rev.2%20Arbitration%20Rules_EN.pdf; Patent Mediation & Arbitration Ctr. of the Unified Patent Court, Mediation Rules arts. 1, 26-27 (adopted Apr. 24, 2026), https://www.pmac-upc.org/sites/default/files/upc_documents/D_AC_06_04112025_rev.2%20Mediation%20Rules_EN.pdf.
• 17Convention on the Recognition and Enforcement of Foreign Arbitral Awards art. III, June 10, 1958, 330 U.N.T.S. 38.
• 18C. civ. arts. 1104, 1194, 1205-1209 (Fr.); Unwired Planet Int'l Ltd. v. Huawei Techs. Co., [2017] EWHC 711 (Pat), [139]-[146] (Eng.); Unwired Planet, [2020] UKSC 37, [8], [12]-[14].