Chapter 7: RAND and Royalty-Bearing Standards

RAND — Reasonable and Non-Discriminatory — is the patent policy model that allows patent holders to charge royalties for their necessary claims while constraining how much they can charge and on what terms. It is the dominant model in industries where companies have invested heavily in patented technology and expect to recover that investment through licensing: wireless communications, video codecs, cellular networks.

The goal, as discussed in Chapter 6, is to restore the results of competition. Even though a standard eliminates competing approaches, the RAND commitment requires the patent holder to behave as if alternatives still existed — licensing at market rates rather than monopoly rates. The patent holder can monetize their investment. The implementer gets access on reasonable terms. That's the theory.

Proponents of RAND argue that it's necessary to incentivize innovation — that companies won't invest in developing new technology for standards unless they can monetize the resulting patents. There is no empirical proof of this. The web was built almost entirely on royalty-free standards, and it's difficult to argue that the web lacks innovation because patent holders weren't collecting royalties on HTTP, HTML, or CSS. The reality is that RAND and royalty-free reflect different business models, not different levels of innovation. Some industries have established patent licensing as a core revenue stream. Others haven't. The choice of patent policy follows from that commercial reality.

This chapter covers how RAND works in practice, how disclosure obligations operate, how disputes over rates play out, and the contentious question of whether RAND-committed patent holders can seek injunctions.

7.1 When Organizations Choose RAND

The choice between royalty-free and RAND is one of the first and most consequential decisions in setting up a standards body.

Royalty-free works when the participants are primarily interested in broad adoption and commodity interoperability. The web standards world is a natural fit — everyone benefits from ubiquitous implementation, and the value is in the products and services built on top of the standard, not in the standard itself.

RAND works when some participants hold valuable patent portfolios developed through significant R&D investment, and they expect to license those patents as part of their business model. The telecom industry is the canonical example. Companies spend years developing wireless technology, file patents on it, and then contribute it to standards bodies with the expectation that they'll license those patents to implementers.

It's worth noting that "RAND" doesn't mean everyone charges royalties. A RAND commitment means you can charge, not that you must. The vast majority of RAND-committed patent holders never set up active licensing programs. They make the commitment, participate in the work, and never seek royalties. Active licensing is concentrated in a few specific industries.

7.2 FRAND: Fair, Reasonable, and Non-Discriminatory

In Europe, the commitment is typically called FRAND — with the F standing for Fair. RAND and FRAND mean the same thing in practice, and no, the F does not mean Free. This is a common misunderstanding that can be consequential.

A RAND or FRAND commitment is a promise to offer a license to necessary claims on reasonable and non-discriminatory terms, negotiated bilaterally between patent holder and implementer outside the standards body. As discussed in Chapter 3, this is a commitment to license — not a license itself. The standards organization facilitates the commitment but doesn't set rates, administer licenses, or broker negotiations.

What Does "Reasonable" Mean?

This is where all the litigation happens. The patent policy terms themselves are rarely contested in court. It's the rates that get fought over, and those fights can involve hundreds of millions of dollars.

"Reasonable" is generally understood to mean the rate the patent holder could command in a hypothetical negotiation where competitive alternatives exist — even though, in reality, the standard has eliminated alternatives. The idea is to strip out the monopoly premium created by standardization and arrive at the rate the patent would command on its own merits.

Courts have developed various methodologies for calculating RAND rates — comparable license analysis, top-down approaches based on the aggregate royalty burden for the standard, bottom-up approaches based on the patent's incremental value. The details are the domain of licensing litigators and economists and are beyond the scope of this book. What matters for practitioners working on the policy side is understanding that "reasonable" is a contested concept, that rates are negotiated or litigated outside the standards body, and that the policy creates the framework but not the answer.

What Does "Non-Discriminatory" Mean?

The non-discriminatory prong requires that the patent holder offer the same (or comparable) terms to similarly situated licensees. You can't charge one implementer a penny and another a dollar for the same technology without justification.

In practice, "non-discriminatory" doesn't mean identical terms for everyone. Volume discounts, cross-licensing arrangements, and differences in the licensee's use case can all justify different terms. What it prohibits is using the licensing process to pick winners and losers among implementers — charging a competitor more simply because they compete with you.

7.3 The Disclosure Obligation

RAND policies typically include a disclosure obligation: if you're aware of patents that may read on the standard, you must tell the group. This is distinct from the royalty-free model, where the focus is on exclusions. In RAND regimes, the focus is on transparency.

How Disclosure Works

Disclosure obligations come in the same two forms as exclusion mechanisms: a formal call for patents tied to a specific milestone in the specification process, or a rolling obligation to disclose whenever you become aware of a relevant patent.

The arguments for and against each approach parallel the exclusion discussion in Chapter 8. A formal call gives participants a concrete deadline and a trigger for internal review. A rolling obligation is self-executing and doesn't depend on the organization following its own procedures. For large organizations that may not have day-to-day visibility into what their standards participants are doing, the formal call is generally more effective — it produces a piece of paper that gets routed to the patent team.

The Standard of Knowledge

Most disclosure policies are based on the actual personal knowledge of the working group representative — not the collective knowledge of the entire company. You don't have to conduct a patent search. You don't have to consult every patent attorney in the building. If the individual who participates in the working group is personally aware of a potentially relevant patent, they must disclose it.

This standard is deliberately narrow. The rationale is that imposing a broader due diligence obligation would make participation too expensive and would deter companies from engaging. The tradeoff is that relevant patents can go undisclosed when the representative genuinely doesn't know about them — even if someone else in the organization does.

The narrow standard creates an obvious incentive problem: send someone who doesn't know your patent portfolio, and you can avoid disclosure. This strategy tends to cancel itself out in practice, because a company that fails to disclose and later tries to assert faces arguments about patent misuse, unclean hands, and — as one court found — antitrust liability.

The Qualcomm v. Broadcom Lesson

The most cited case on disclosure obligations is Broadcom Corp. v. Qualcomm Inc. Qualcomm participated in a standards body, failed to disclose patents it held on the standard, and later asserted those patents against implementers. The litigation was compounded by discovery misconduct — Qualcomm produced critical documents only on the eve of trial.

The court found that Qualcomm's failure to disclose constituted anticompetitive behavior. But the reasoning went further than the patent policy's own terms. Rather than applying the policy's disclosure standard — actual personal knowledge of the representative — the court articulated a broader standard based on the "reasonable expectations" of the other participants. This effectively created a disclosure obligation beyond what the policy's text required.

The case is important but problematic. It was decided against a backdrop of egregious misconduct — the court was looking for a way to sanction Qualcomm's behavior, and the disclosure failure provided a vehicle. Whether the "reasonable expectations" standard would be applied to a good-faith participant who simply didn't know about a relevant patent is unclear. The case hasn't been widely followed, but it remains a cautionary example of what happens when disclosure goes wrong.

No Duty to Search — But Disclosure Is Not Optional

Almost universally, RAND policies do not impose an obligation to conduct a patent search. The actual personal knowledge standard is the floor and the ceiling. This is worth emphasizing because participants — especially engineers — sometimes assume they need to do more. They don't. And the organizations themselves shouldn't be conducting searches on behalf of participants.

What does need to be emphasized — because this is a point many participants get wrong — is that disclosure itself is mandatory, not optional. If you have a call for patents and you have a patent that you're aware of, you must disclose it. This is not a discretionary decision. It is a requirement of the policy.

Engineers and business teams sometimes resist disclosure because they worry it will look bad, send a negative signal to the community, or derail the work. Those concerns are understandable but irrelevant. If the policy requires disclosure based on actual personal knowledge, and the representative knows about a relevant patent, there is no choice to make. The obligation is clear.

The failure to disclose a known patent doesn't eliminate the licensing obligation — the RAND commitment still applies whether or not you disclosed. But non-disclosure creates significant enforcement problems if you later try to assert. It can be characterized as patent ambush, it invites antitrust scrutiny, and it undermines your credibility in the community. Disclosure may feel uncomfortable, but the alternative is worse.

7.4 The Injunction Question

One of the most contested issues in standards patent law is whether a holder of a standard-essential patent who has made a RAND commitment can seek an injunction against an implementer. The arguments on each side are substantial.

The Case Against Injunctions

If a patent holder has committed to license on reasonable and non-discriminatory terms, seeking an injunction — an order that would prevent the implementer from shipping their product — is arguably inconsistent with that commitment. The patent holder has already agreed to license. The only open question is the rate. An injunction in this context becomes a tool for extracting above-market royalties under the threat of shutting down the implementer's business. This is the patent holdup problem: the patent holder uses the threat of an injunction to force a settlement far above what a reasonable royalty would be.

The Case For Injunctions

Patent holders argue that without the possibility of an injunction, implementers have no incentive to negotiate in good faith. If the worst outcome for an implementer is paying a reasonable royalty — the same amount they'd pay if they had negotiated up front — there's no cost to refusing to negotiate and simply implementing without a license. The implementer captures the full value of the standard while the patent holder bears the cost of enforcement. This is the implementer holdout problem: the implementer delays, ignores licensing overtures, and forces the patent holder to litigate just to get paid what was already owed.

How Courts Have Addressed It

Courts in different jurisdictions have taken different approaches, and the law is not settled.

In the United States, the Supreme Court's 2006 decision in eBay Inc. v. MercExchange established that injunctions in patent cases are not automatic. Courts must apply a four-factor equitable test: irreparable injury, inadequacy of monetary remedies, balance of hardships, and the public interest. In the SEP context, U.S. courts have generally been reluctant to grant injunctions where the patent holder has made a FRAND commitment, reasoning that monetary damages — a reasonable royalty — are an adequate remedy. The Ninth Circuit reinforced this in Microsoft Corp. v. Motorola Inc., finding that a demand for an injunction on FRAND-committed patents was inconsistent with the licensing obligation.

In Europe, the Court of Justice of the European Union took a different approach in Huawei Technologies v. ZTE Corp. (2015). The CJEU established a structured negotiation framework: before seeking an injunction, the SEP holder must notify the implementer and make a specific FRAND licensing offer. The implementer must respond diligently and in good faith. If the implementer refuses to engage — if they are an "unwilling licensee" — the patent holder may seek an injunction without running afoul of EU competition law. This framework gives both sides procedural obligations and reserves injunctions for cases where the implementer is genuinely refusing to negotiate.

The practical result is that injunctions on FRAND-committed patents are rare in the United States and available but procedurally constrained in Europe. Other jurisdictions — China, the UK, India — are developing their own approaches, and the interaction of multiple jurisdictions in global licensing disputes adds further complexity.

Why This Matters for Policy Design

For practitioners drafting or evaluating patent policies, the injunction question is relevant even though it plays out in courts rather than in the standards body itself. A policy that is explicitly silent on injunctions leaves the question to the courts. Some policies include language that the commitment is to license on RAND terms "without seeking injunctive relief," though the enforceability of such provisions varies. Others deliberately leave it open, taking the position that injunctive relief is a judicial question, not a contractual one.

The injunction debate is ultimately about the balance of power between patent holders and implementers — and that balance differs depending on the industry, the jurisdiction, and the specific facts. Detailed treatment of post-implementation licensing disputes is beyond the scope of this book, but understanding the injunction dynamic is essential context for the patent policy choices covered here and in Chapter 8.

7.5 Patent Pools and Collective Licensing

In industries where a standard is covered by patents from many holders — video codecs and wireless are the primary examples — patent pools often form to simplify licensing.

A patent pool aggregates patents from multiple holders into a single licensing program. Instead of negotiating separate licenses with dozens of patent holders, an implementer takes one license from the pool and gets access to all the pooled patents. The pool administrator collects royalties and distributes them to the patent holders according to an agreed formula.

Patent pools are generally formed outside the standards body, often by a licensing administrator like MPEG LA or Via Licensing. The pools typically employ an independent expert who evaluates whether each submitted patent is truly a necessary claim before it's included in the pool. This gatekeeper function is important — without it, patent holders could submit marginal patents and dilute the pool's value.

One thing to watch: the independent evaluators are paid by the pool and tend to be permissive in what they consider essential. More patents in the pool means more revenue for the pool and its administrator. This doesn't mean the evaluation is corrupt, but the incentive structure is worth understanding.

Patent pools raise antitrust questions — competitors collectively setting a price for their patents looks like price-fixing. Courts and regulators have generally approved patent pools for standards when they include only necessary claims, when they are non-exclusive (patent holders can still license independently), and when they don't extend to non-essential patents. The key distinction is that pooling necessary claims for a standard reduces transaction costs without restricting competition, while pooling broader patent rights could suppress it.

For practitioners, the existence of a patent pool may be the first concrete answer to a client's question: "what will it cost to implement this standard?" If a pool exists, it provides a published rate. If it doesn't, the answer is the familiar: implement and be prepared for someone to knock on your door.

7.6 Over-Disclosure and Strategic Behavior

In RAND regimes, a significant practical problem is over-disclosure — patent holders disclosing far more patents than are actually essential.

Because disclosure obligations are based on a reasonable belief that a patent might be a necessary claim — not a formal determination — patent holders have every incentive to disclose broadly. If you're planning to license, disclosing early establishes your position. If you're not sure, it's safer to disclose than to risk the consequences of non-disclosure. Note that this is a RAND-specific issue. In royalty-free organizations, there is generally no need for patent disclosure since participants are getting a royalty-free license regardless — the relevant mechanism there is the exclusion process, not disclosure.

Over-disclosure creates real problems for the standards development process. A working group confronted with a large volume of patent disclosures may feel compelled to evaluate them, potentially triggering design-around efforts, slowing the development cycle, or diverting engineering resources from technical work to patent analysis. In some cases, over-disclosure is strategic — a way to signal the size of a patent position and influence the technical direction of the spec.

The result is that the disclosure list for a mature RAND standard can be enormous, with far more patents listed than are actually essential. Implementers looking at the disclosure list may overestimate their royalty exposure. Patent holders who later form pools will submit patents that were disclosed but may not survive independent evaluation.

For the standards body, managing the volume of disclosures is an ongoing challenge. For implementers trying to assess their exposure, over-disclosure creates noise that makes due diligence more difficult.

7.7 Standard-Essential Patents and the Cross-Licensing Reality

A term you'll encounter frequently in RAND discussions is SEP — standard-essential patent. A SEP is simply a patent that contains one or more claims that are necessary to implement a standard. It's the patent-level equivalent of the "necessary claims" concept discussed in Chapter 6. If a patent has at least one claim that is essential to a standard, the entire patent is often referred to as a SEP, even though not all of its claims may be standard-essential.

SEPs matter because they are the patents subject to RAND commitments. They are the patents that get disclosed, that get submitted to patent pools, and that are the subject of licensing negotiations and litigation. The term is used pervasively in the wireless, video codec, and telecom industries, and increasingly in policy discussions by regulators and legislatures.

How Licensing Actually Works

The textbook model of RAND licensing — patent holder offers a license, implementer negotiates terms, they reach agreement on a per-standard royalty — is an oversimplification of how licensing works among large companies.

In practice, SEP licensing between major companies is rarely done on a standard-by-standard basis. Instead, it happens as part of broad cross-licensing agreements that cover both standard-essential and non-essential patents across multiple technology areas. Company A licenses its entire portfolio to Company B, and Company B does the same in return. The agreement may cover SEPs for wireless standards, SEPs for video codecs, non-essential patents on implementation techniques, and patents entirely unrelated to standards — all in a single deal.

These cross-licenses are the dominant form of patent licensing among large technology companies. They are negotiated bilaterally, are almost always confidential, and often involve balancing payments where one party's portfolio is larger or more valuable than the other's. The RAND commitment provides the floor — neither party can refuse to license SEPs on reasonable terms — but the actual deal encompasses far more than SEPs alone.

Patent pools and standalone SEP licensing programs do exist and are significant in specific areas — particularly wireless (where pool-like structures like Avanci operate) and video codecs (where MPEG LA and Via Licensing administer pools). But for the largest companies, these are often supplementary to the broader cross-license relationships.

This distinction matters for practitioners because it means that the RAND commitment, while foundational, is one input into a much larger licensing dynamic. A client asking "what will it cost to license the SEPs for this standard?" may be asking the wrong question. The more relevant question is often: "what is the state of our overall patent relationship with the companies that hold SEPs in this space?"

7.8 RAND-RF Hybrids and Emerging Models

RAND-Z with RAND fallback is a model where participants initially commit to license at zero royalty, but retain the right to fall back to RAND (royalty-bearing) terms if they choose. In theory, this offers the best of both worlds — royalty-free as the default, with a safety valve for patent holders who need to monetize. In practice, these structures tend to produce a race to the bottom. The moment one participant elects the RAND fallback and begins charging, other patent holders follow — because there's no incentive to continue offering royalty-free terms when your competitors are collecting royalties on the same standard. The RAND-Z default effectively collapses into a RAND regime once anyone exercises the fallback.

Dual-mode organizations like JDF allow different working groups within the same organization to operate under different IPR modes. One working group might use royalty-free while another uses RAND, depending on the commercial dynamics of the technology and the participants involved. This flexibility avoids forcing a one-size-fits-all choice at the organizational level.

These hybrid approaches reflect a practical reality: the binary choice between RAND and royalty-free doesn't always map cleanly to the needs of a particular project. The trend is toward more flexibility in policy selection, allowing the IP terms to match the specific context of the work rather than the institutional default of the hosting organization.

7.9 The Practical Reality

For all the complexity of RAND policy terms, the practical reality for most implementers is straightforward: implement the standard and see what happens.

In the vast majority of cases, nobody comes knocking. Active licensing programs are concentrated in wireless, video codecs, and a handful of other areas. Outside those spaces, RAND commitments are largely theoretical — they exist in the policy, patents may be disclosed, but no one sets up a licensing program and no money changes hands.

When licensing does happen, it happens outside the standards body through bilateral negotiation. If the parties can't agree, the dispute goes to court — not back to the standards organization. The standards body's role ends with the commitment. Everything after that is between the patent holder, the implementer, and potentially a judge.

The most difficult conversation with a client in this space is the one where they ask: "How much will it cost to implement this standard?" In most cases, the honest answer is: we don't know. It could cost nothing. It could cost a meaningful amount. The uncertainty is inherent in the RAND model, and no amount of policy drafting eliminates it. What the policy does is ensure that if someone does come knocking, there's a framework — reasonable terms, non-discriminatory treatment — that constrains the negotiation. The alternative would be no framework at all, which would be worse.