Abstract

Standard Essential Patents (SEPs) are patents that protect technology which must be used to comply with a technical standard. Cellular standards such as 4G and 5G make it possible for devices made by different manufacturers to communicate with one another. The same principle is increasingly important to connected vehicles, smart appliances, industrial sensors, logistics equipment and other products forming part of the Internet of Things (IoT). As these sectors adopt cellular connectivity, companies that traditionally operated outside the telecommunications industry may find themselves facing licensing demands and patent litigation from owners of large telecommunications portfolios.

The economic purpose of standardisation is to create interoperability and expand the market for compatible products. It also creates a particular form of patent leverage. Before a standard is adopted, manufacturers may be able to choose between competing technologies. Once a technology becomes embedded in a widely used standard, changing to an alternative may be commercially unrealistic. An implementer may therefore have to use the patented technology to make a standards-compliant product. This can give the patent owner bargaining power beyond the value that the invention would have commanded before standardisation.

1. Introduction

The concern is commonly described as patent hold-up. A SEP owner may seek royalties that reflect the implementer's dependence on the standard, the cost of switching or the value of the finished product, rather than the patented invention's incremental technological contribution. The opposite concern is implementer hold-out: a manufacturer may continue using patented technology while delaying negotiations, withholding sales information or making unreasonable counter-offers. If hold-out is tolerated, innovators may be deprived of a fair return on research and development. The legal challenge is to prevent both forms of opportunism.

Standard-setting organisations address this tension through intellectual property policies. ETSI's IPR Policy, particularly Annex 6, requires members to disclose relevant intellectual property and provides for an irrevocable undertaking to license essential patents on fair, reasonable and non-discriminatory terms, commonly called FRAND. FRAND is not a single universal price. It is a commitment to negotiate terms that are fair in the circumstances, reasonable in light of the patented contribution and non-discriminatory between similarly situated licensees. ETSI's policy does not itself prescribe one mandatory licensing level or royalty formula.

A declaration that a patent is essential does not prove that it is valid, infringed or necessarily essential. A court must examine the patent claims, the technical standard and the accused product. The distinction is important because SEP declarations are often made by patent owners and may cover large portfolios. The existence of a declaration should be treated as relevant information, not as conclusive evidence. The same care is needed when a portfolio is licensed globally: the number of declared patents is not a substitute for evidence about essentiality, validity, geographic coverage and actual technological contribution.

2. The Appropriate Level of SEP Licensing

A central dispute concerns the appropriate point in the supply chain at which a SEP should be licensed. Under component-level licensing, the patent owner licenses the supplier of the component that implements the standard, such as a modem, chipset or cellular communication module. Under end-device licensing, the licence is taken by the manufacturer of the finished product, such as a vehicle maker or smart-appliance company. Each approach has commercial advantages. Component-level licensing may connect the royalty more closely to the technology that practices the standard and may allow a supplier serving several manufacturers to obtain one licence. It can also reduce repeated negotiations and uncertainty for smaller IoT businesses.

End-device licensing may be attractive because the original equipment manufacturer (OEM) controls the finished product, sales information and commercial relationship with customers. A single licence to the OEM may cover multiple component suppliers and avoid disputes about downstream use.

Patent owners also argue that the finished product is the natural unit for portfolio licensing, particularly where a device incorporates several technologies and the value of connectivity is not confined to one chip. However, a royalty calculated on the entire product price may overstate the contribution of a small connectivity component, especially in vehicles and industrial equipment.

In FTC v Qualcomm Inc., 969 F.3d 974 (9th Cir. 2020), the United States Court of Appeals for the Ninth Circuit reversed the district court's antitrust judgment against Qualcomm. It held that the FTC had not established the alleged antitrust violations on the record before it and emphasised that antitrust law does not generally impose a duty to license rivals or deal with every requesting party. The judgment is important to the component-level debate, but it does not create a universal rule that SEP owners may refuse component licences in every jurisdiction. Contractual FRAND commitments and competition laws  outside the United States may lead to different outcomes.

The Ninth Circuit's reasoning was confined to the FTC's pleaded Sherman Act theory and the evidence in that record. It distinguished a possible breach of a FRAND undertaking from the separate question whether conduct satisfies the elements of federal antitrust law. The court found no sufficient showing that Qualcomm's licensing practices harmed competition in the relevant markets. The judgment therefore informs the US component-licensing debate, but it does not settle contractual FRAND duties under every standard-setting policy or the law of other countries.

The automotive supply chain was directly implicated in Continental Automotive Systems, Inc. v Avanci, LLC, No. 20-11032 (5th Cir. 2022). Continental, an automotive supplier, challenged a licensing model under which SEP holders licensed vehicle manufacturers rather than component suppliers. The Fifth Circuit affirmed dismissal of the antitrust claims, finding the pleaded theory insufficient under US law. The opinion was unpublished and non-precedential.It nevertheless illustrates the practical tension: a module supplier may provide the connectivity technology, while the vehicle manufacturer controls the final product and its global sales.

Continental argued that a direct component-level licence was necessary for suppliers of cellular modules. The Fifth Circuit held that the complaint did not adequately allege an antitrust injury caused by the refusal to license component suppliers. The opinion was unpublished and non-precedential. Its significance is consequently limited: it rejects the particular antitrust claim pleaded, rather than declaring that OEM-level licensing is compulsory or that a FRAND undertaking can never require a component-level offer. Neither component-level nor OEM-level licensing should be treated as automatically correct. A sound approach is to require SEP owners to make FRAND licences genuinely accessible to willing implementers at a technically and commercially appropriate level. The parties should be able to agree on a single effective licence with clear terms addressing downstream coverage, exhaustion, indemnity, audit rights and prevention of double recovery. Courts should consider the actual patent claims, the component's role, the parties' licensing history and the risk that more than one party in the supply chain will be charged for the same use.

3. SEP Injunctions and Abuse of Dominance

The most influential European decision on SEP injunctions is Huawei Technologies Co. Ltd v ZTE Corp., Case C-170/13, EU:C:2015:477. The Court of Justice of the European Union recognised that a SEP owner which has given a FRAND undertaking may hold a dominant position and that seeking an injunction may, in particular circumstances, amount to abuse under Article 102 of the Treaty on the Functioning of the European Union. The Court did not prohibit injunctions altogether. Instead, it set out a negotiation sequence designed to protect a willing implementer from coercive demands while preserving the patent owner's ability to act against an unwilling licensee.

The Court's framework requires the SEP owner first to give notice identifying the patent and the way it is said to be infringed. If the implementer indicates willingness to take a FRAND licence, the owner must make a specific written offer, including the royalty and its calculation. The implementer must respond diligently and in good faith; if it rejects the offer, it should make a specific written counter-offer. If it continues using the patent after rejection of that counter-offer, it should provide appropriate security and render an account. The Court also recognised that the parties may agree to have the royalty fixed by an independent third party. The sequence is intended to distinguish a genuine licensee from a party using negotiations merely to delay payment.

Before seeking an injunction or product recall, the SEP owner should alert the alleged infringer by identifying the patent and explaining how it is infringed. If the implementer expresses willingness to conclude a FRAND licence, the patent owner should make a specific written offer setting out the royalty and its calculation. If the offer is rejected, the implementer should promptly make a specific written counter-offer on FRAND terms. Where the implemented continues using the patent after its counter-offer is rejected, it should provide appropriate security and account for past use. The parties may also agree to have the royalty determined by an independent third party.

The principle is that an injunction should not be used as leverage to force a willing business to accept terms that are not FRAND. At the same time, an implementer cannot avoid payment merely by asserting that it is willing to negotiate. A party that ignores notices, delays without justification, makes unserious counter-offers, refuses reasonable security or conceals sales information may be treated as unwilling in substance. The court must examine the conduct of both parties throughout the negotiation, rather than relying on a label used by either side.

The German Federal Court of Justice applied the European framework in Sisvel International S.A. v Haier, KZR 36/17 (5 May 2020). The court stressed that an implementer must demonstrate genuine willingness to take a FRAND licence and cannot simply postpone meaningful engagement until every patent issue has been finally resolved. The judgment is often cited in discussions of hold-out. It does not, however, release the SEP owner from the obligation to make a sufficiently reasoned FRAND offer. A global portfolio proposal may be commercially legitimate, but it is not automatically FRAND merely because it is global.

The BGH assessed the parties' conduct as a whole and stressed that willingness must be demonstrated by concrete and timely steps, not simply asserted. An implementer cannot insist that all validity and infringement disputes be finally resolved before engaging with a licence proposal. At the same time, the decision does not make the SEP owner's offer automatically FRAND: the owner must still present a sufficiently specific and reasoned proposal, and the implementer remains entitled to challenge the terms in good faith.

In the United States, patent injunctions are governed by the equitable test in eBay Inc. v MercExchange, L.L.C., 547 U.S. 388 (2006), which requires proof of irreparable injury, inadequacy of monetary remedies, a balance of hardships favouring relief and consistency with the public interest. This makes an injunction a discretionary remedy rather than an automatic consequence of infringement. US courts have also been cautious about treating a FRAND disagreement as an antitrust violation without proof of the elements of the relevant statute. The European and American approaches therefore differ in legal method, even though both seek to prevent misuse of market power and preserve legitimate patent enforcement. The four equitable factors are irreparable injury, inadequacy of remedies at law, the balance of hardships and the public interest. In an SEP dispute, the availability of a monetary royalty or security may be relevant, as may the consequences of stopping a product line. This is a US patent-remedy test and should not be confused with the EU competition-law analysis in Huawei v ZTE.

4. The Indian Legal Position

Indian law must reconcile the exclusive rights of a patentee with the public interest in competition and access to technology. Section 48 of the Patents Act, 1970 sets out the exclusive rights of a patentee, subject to the Act. Section 84 permits compulsory licensing after three years from the grant of a patent where statutory grounds are established, including failure to satisfy the reasonable requirements of the public, lack of availability at a reasonably affordable price or non-working of the patent in India. Sections 3 and 4 of the Competition Act, 2002 address anti-competitive agreements and abuse of dominant position. The interaction between these statutes has been contested in SEP litigation.

In Telefonaktiebolaget LM Ericsson v Competition Commission of India and connected appeals, the Delhi High Court's Division Bench delivered judgment on 13 July 2023 on the relationship between the Patents Act and the Competition Act. The litigation raised the question whether the Competition Commission of India could investigate allegations concerning SEP licensing practices while patent proceedings were pending. The judgment is significant to the institutional relationship between patent enforcement and competition scrutiny. It should not be reduced to a general proposition that every SEP licensing practice is immune from competition law; the precise holdings and subsequent procedural history must be checked before relying on it.

In Intex Technologies (India) Ltd v Telefonaktiebolaget LM Ericsson, 2023:DHC:2243-DB (29 March 2023), the Delhi High Court considered essentiality, infringement, FRAND conduct and interim injunctive relief. The court examined whether the implementer had behaved as a willing licensee and made clear that resort to the Competition Commission, by itself, does not necessarily establish unwillingness. The substance of the negotiations remains important. The case is useful in showing that an implementer must engage meaningfully, but also that a SEP owner must satisfy its own FRAND obligations.

The court considered the parties' licensing exchanges and the implementer's response to Ericsson's offer. It treated willingness as a matter of conduct and negotiation history, rather than assuming that filing a complaint before the Competition Commission automatically made the implementer unwilling. The case is useful for the proposition that both sides must participate meaningfully: the implementer should make reasoned responses, while the SEP owner must put forward a genuine FRAND proposal.

In Telefonaktiebolaget LM Ericsson v Lava International Ltd., 2024:DHC:2698 (28 March 2024), the Delhi High Court delivered a detailed final judgment concerning patent validity, essentiality, infringement, FRAND royalties and damages in a mobile-phone dispute. The judgment demonstrates the importance of technical evidence linking patent claims to the relevant standard and accused products. It also provides an important Indian example of judicial examination of royalty evidence and licensing comparables. A declared SEP should not be treated as automatically valid or infringed; the patentee must establish the relevant legal and technical elements.

The court examined the asserted patent claims against the relevant standard and the accused products, alongside validity and infringement evidence. It then considered royalty and damages evidence in the context of the dispute. The judgment is important because it illustrates that a declaration of essentiality does not replace proof in court; the patentee must establish the relevant technical and legal elements. It is also a substantial Indian example of judicial assessment of FRAND-related monetary relief.

For India, neither categorical immunity for SEP owners nor automatic condemnation of injunction applications is justified. Courts should require proof of essentiality and infringement, meaningful notice, a transparent FRAND offer and genuine negotiation. Injunctive relief should be proportionate to the circumstances, especially where a monetary remedy or security can protect the patent owner without unnecessarily excluding connected products from the market. Competition scrutiny should focus on exclusionary conduct and market effects, rather than treating the mere exercise of a patent right as proof of abuse.

5. Global FRAND Licensing and Judicial Jurisdiction

Global licensing presents a further difficulty. Patents are territorial rights: an Indian patent is governed by Indian law, and a UK patent by UK law. Yet SEP portfolios and product markets are international. A vehicle or connected device may be manufactured in one country, incorporate components from another and be sold across many jurisdictions. Separate negotiations for every national patent can be expensive and inefficient. For that reason, parties frequently seek a single worldwide portfolio licence. The legal question is whether a national court may determine global contractual terms without claiming authority to decide the validity or infringement of every foreign patent. The leading decision is Unwired Planet International Ltd v Huawei Technologies (UK) Co Ltd, [2020] UKSC 37. The UK Supreme Court upheld the

English court's ability, in the circumstances of the case, to determine the terms of a worldwide FRAND licence and to grant an injunction in respect of UK patent infringement unless Huawei entered into that licence. The Court considered the commercial reality of global telecommunications portfolios and treated a worldwide licence as capable of being FRAND. It also held that the non-discrimination obligation does not necessarily require every licensee to receive the lowest rate ever agreed with another licensee. Differences may be justified by relevant commercial circumstances.

The Supreme Court's reasoning did not give English courts authority to determine the validity of foreign patents. It treated the worldwide licence as a contractual FRAND outcome between the parties in the circumstances of the case, while the injunction related to infringement of UK patents. The Court also explained that the non-discrimination limb does not invariably require the same rate for every licensee or an automatic entitlement to the lowest rate ever granted; differences may be justified by relevant commercial circumstances.

The decision promotes efficiency, but it raises questions of jurisdiction and comity. A global rate fixed by one court may affect parties and patent rights in countries that did not participate in the proceedings. A defendant may feel compelled to accept a worldwide licence to avoid an injunction on a domestic patent. This can encourage forum shopping and create conflict where courts in different countries reach different conclusions. A court determining global contractual terms should therefore identify its jurisdictional basis, distinguish rate-setting from foreign patent validity, consider parallel proceedings and avoid orders that interfere unnecessarily with other courts.

In TCL Communication Technology Holdings, Inc. v Telefonaktiebolaget LM Ericsson, 943 F.3d 1360 (Fed. Cir. 2019), the US Court of Appeals for the Federal Circuit vacated and remanded aspects of a district court's global FRAND decision, including issues concerning the right to a jury trial and the treatment of foreign patent matters. The case illustrates the procedural difficulty of global rate-setting. In China, decisions and proceedings including Huawei v Conversant and Xiaomi v InterDigital have brought anti-suit relief and competing jurisdictional claims into focus. These disputes reinforce the need for judicial restraint and respect for territorial patent rights.

The case demonstrates that global rate-setting raises procedural questions as well as economic ones. The Federal Circuit addressed the jury-trial right and the treatment of foreign patent issues, and remanded aspects of the district court's decision. It cautions against assuming that a court can resolve every issue in a worldwide portfolio dispute without respecting the procedural protections and territorial character of patent rights.

There is no universally accepted mathematical formula for a FRAND rate. Courts commonly examine comparable licences, portfolio strength, geographic coverage, essentiality, validity risk, product scope and the value attributable to the patented technology. A top-down approach may estimate an aggregate royalty burden for the standard and allocate a proportionate share to a particular portfolio. An incremental-value approach asks what value the patented invention adds over the next-best alternative before standardisation. Neither method is conclusive in every case; the reliability of the evidence and the assumptions used matter more than the label attached to the method.

6. Patent Pools, IoT and the Way Forward

Comparable licences can provide useful evidence, but they must be adjusted for differences in portfolio size, geographic coverage, duration, volume, cross-licensing, litigation risk and the products covered. A settlement reached under the pressure of litigation may not be a perfect market benchmark. A rate expressed as a percentage of the finished product's price may overstate the contribution of a connectivity component in a complex product. A per-unit royalty may be more predictable, but may fail to reflect differences in product functionality or licence scope. Courts should explain why the selected royalty base and rate, considered together, produce a proportionate result.

Patent pools can reduce transaction costs by aggregating patents held by multiple owners and offering a package licence. They may be useful in IoT markets where a manufacturer needs access to numerous cellular patent families. However, a pool is not automatically pro-competitive. Its effects depend on whether the included patents are genuinely essential and complementary, whether participation is voluntary, whether independent licensing remains possible, how royalties are allocated and whether the pool excludes competing technologies. Competition authorities should examine the pool's structure and effects rather than treating aggregation as inherently lawful or unlawful.

The growth of connected vehicles, smart appliances and industrial equipment makes licensing predictability especially important. These products differ from smartphones in price, margins, production volumes, expected service life and the role played by connectivity. A model designed for high-volume consumer devices may not suit a low-volume industrial sensor or a vehicle expected to remain in service for many years. A practical licence may need to address different product categories, territories, supply-chain participants and lifecycle stages. Small and medium-sized IoT businesses should have a realistic route to obtain licences without facing disproportionate transaction costs.

The European Commission proposed a Regulation on standard essential patents in April 2023, which contemplated an EUIPO competence centre, an SEP register, essentiality checks and a process for FRAND determination. The proposal sought to improve transparency and reduce information asymmetry, particularly for IoT implementers. It did not become an operative regulation: the European Parliament's Legislative Observatory records the Commission's withdrawal of the proposal in October 2025 and subsequent parliamentary action concerning that withdrawal. The proposal is therefore relevant as a policy document, not as binding law.

A balanced framework should require SEP owners to give specific notice, identify relevant patents and explain the royalty methodology. Implementers should respond promptly, provide relevant sales information, make reasoned counter-offers and provide appropriate security where continued use is not disputed. Courts should insist on reliable technical proof, evaluate both parties' conduct and use injunctions proportionately. For global licences, they should rely on transparent evidence, explain the jurisdictional basis for their decision and respect parallel proceedings. Patent pools should disclose their essentiality methods and royalty allocation principles, while preserving meaningful opportunities for independent licensing. In India, specialist judicial capacity in patent economics and standards engineering would improve the quality of SEP decisions. Coordination between the Patents Act and Competition Act should be clear enough to prevent either regime from being used to frustrate the other. Courts should protect legitimate patent enforcement while scrutinising conduct that exploits standard-created market power. A predictable FRAND framework can encourage investment in 5G, future 6G and connected technologies without turning connectivity into an unpredictable tax on every product that uses a standard.

7. Critical Analysis and Author's Perspective

The position advanced in this article is that FRAND must be understood as a two-sided commitment. A SEP owner should receive a reasonable return for genuine technological contribution, investment and risk. At the same time, standardisation should not permit the owner to charge for value created by the standard itself, the implementer's brand, network effects or the cost of abandoning an established standard. The implementer, for its part, should not treat the FRAND undertaking as a licence to use patented technology indefinitely without payment. The law should therefore discourage both patent hold-up and implementer hold-out, rather than favouring one side categorically.

In my view, licensing should be flexible as to the level of the supply chain, but access to a FRAND licence must be real rather than theoretical. Where a component supplier independently implements the standard and seeks a licence for that component, the SEP owner should give serious consideration to licensing at that level. This is particularly important in automotive and industrial IoT markets, where a specialist supplier may sell the same communication module to many manufacturers. A direct component licence can improve predictability and reduce repeated negotiations. However, an OEM-level licence may be commercially reasonable where the parties agree that the OEM is the appropriate licensee and the arrangement prevents double recovery. The decisive considerations should be technical implementation, commercial practice, downstream coverage and the availability of an effective licence—not a rigid rule that always favours the component supplier or the OEM. Injunctions should remain available, but they should be a remedy for genuine unwillingness and proven infringement, not a shortcut for extracting a disputed royalty. Before granting relief that could stop an entire vehicle line or connected-device business, a court should examine whether the patent is valid, essential and infringed; whether the SEP owner made a reasoned FRAND offer; whether the implementer responded promptly with a credible counter-offer; and whether security or monetary relief can protect the patent owner. This approach gives effect to the balance reflected in Huawei v ZTE while recognising the concerns about hold-out identified in Sisvel v Haier.

Global rate-setting requires similar restraint. A worldwide licence may be commercially efficient where the parties operate internationally and the portfolio is genuinely global. Yet a court should clearly distinguish determining contractual FRAND terms between parties from deciding the validity or infringement of foreign national patents. Courts should explain their jurisdictional basis, consider parallel proceedings and avoid unnecessary interference with other jurisdictions. Comparable licences, portfolio strength, geographic coverage, essentiality evidence and the incremental value of the patented technology should be disclosed and tested as far as confidentiality permits.

For India, the strongest path is a coordinated approach under the Patents Act, 1970 and the Competition Act, 2002. Patent enforcement should not be blocked merely because a competition complaint has been filed; equally, the existence of a patent should not automatically immunise conduct that may amount to abuse of dominance. Indian courts and regulators should focus on evidence, market effects, negotiation conduct and proportional remedies. Greater technical expertise, transparent licensing explanations and carefully managed access to comparable agreements would make outcomes more predictable for both innovators and manufacturers.

Ultimately, the objective should be a licensing environment in which innovation is rewarded, access to standards remains practical and the cost of connectivity is not inflated by avoidable bargaining power. This balanced approach is especially important as 5G, future 6G and IoT technologies become embedded in vehicles, factories, public infrastructure and ordinary consumer products.