Drafting Intellectual Property Agreements: Best Practices From a Litigator’s Perspective

9/25/2015 Articles

This article was published in the Summer 2015 edition of the Continuing Education of the Bar's California Business Law Practitioner, Volume 30, Number 3.

By Eugene Y. Mar, Erik C. Olson and Marc Tarlock


As intellectual property licensing continues to grow more prevalent, legal practitioners and business personnel are being asked to craft and negotiate agreements that can significantly impact a business’s ability to compete in a particular field or market. This article collects a number of best practices learned the hard way—through litigation of agreements and arrangements that went awry. Whether negotiating a complex cross-license of intellectual property rights or drafting a supplier agreement, practitioners should consider the license provisions and practice notes discussed below to ensure that a party’s licensing objectives are met.

The first part of this article focuses on the tension between licensors and licensees when it comes to negotiating the scope of an IP license and the protection afforded to business partners, suppliers, and downstream users of a licensed technology. The second part focuses on dispute resolution and indemnity clauses. Although the choices of forum, law, and procedural rules that will govern disputes arising under an IP license may seem insignificant compared with core business terms such as the scope of the license and royalty rates, they have serious ramifications should litigation arise concerning the license. They not only limit how a litigator can vindicate his or her client’s rights under the agreement, but also can have a profound substantive impact on the outcome. Finally, the third part of this article discusses exclusive dealing provisions that can be used in licensing intellectual property to increase market share and stave off competitors. Under a range of circumstances discussed below, licensors can include these types of provisions with only minimal risk.


The core of any IP license is the scope of the license grant. During the course of a license negotiation, parties will inevitably discuss what protections are afforded to downstream customers and users of a technology, business or joint venture partners, and suppliers. A number of approaches can be used to handle these issues, each with its own benefits and drawbacks. Some approaches raise legal issues pertaining to exhaustion or estoppel. Other approaches that focus on limiting the scope of the license face greater scrutiny due to changes in the law over the past few years. Each of these approaches is discussed in greater detail below.

Protections for the Licensee’s Customers, Suppliers, and Business Partners

One of the most common provisions that licensees seek in an IP license is protection for the consumers or end users of the licensee’s technology and the business partners and suppliers for a particular licensee. Downstream consumers enjoy considerable protection under the doctrines of patent exhaustion and first sale, but the same is not true for business partners and suppliers. For the benefit of downstream consumers, the U.S. Supreme Court’s ruling in Quanta Computer, Inc. v LG Electronics, Inc. (2008) 553 US 617, firmly held that “[t]he authorized sale of an article that substantially embodies a patent exhausts the patent holder’s rights and prevents the patent holder from invoking patent law to control postsale use of the article.” 553 US at 638. More recently, the Supreme Court in Kirtsaeng v John Wiley & Sons, Inc. (2013) ___ US ___, 133 S Ct 1351, and the Ninth Circuit in Omega S.A. v Costco Wholesale Corp. (9th Cir 2015) 776 F3d 692, recognized broad exhaustion of copyrights after an authorized first sale, regardless of whether the sale occurred in the United States or overseas. However, business partners and suppliers are involved in a different part of the manufacturing chain. They do not typically sit downstream from the licensee and thus may not be able to argue that the first sale of the licensed technology exhausted the licensor’s rights.

To protect business partners, suppliers, and even downstream users, it may be a good practice for licensees to include a provision that specifically grants a narrow license to end users strictly limited to their use of the licensed product or technology. Licensors should take particular care in drafting these provisions to ensure that the license is strictly limited to the acts of using, supplying, or jointly developing and providing the licensed technology. By narrowly limiting the license to the business partners, suppliers, and downstream users to the specific acts related to providing or using the licensed technology, a licensor minimizes the risk that the license grant will extend to other technologies or products beyond the licensed technology.

Another approach to ensuring protection for the licensee’s business partners, suppliers, and downstream users of the licensed technology is to incorporate a clause that specifies that the licensor will not assert that the licensee’s technology satisfies any particular element of a patent claim or any specific element of a cause of action against any other party (a “no-assertion” clause). For patent licenses, such a clause typically includes language such as the following:

[Licensor] agrees that it will not assert that [Licensee’s technology] directly or indirectly infringes or satisfies any limitation of any claim of the [licensed patents] against any third party.

For other types of IP licenses, the language will specify that the licensee’s technology or product will not be used to satisfy any element of any claim arising in law or equity, cause of action, or demand. Including such clauses will create an estoppel effect against the licensor and likely prevent the licensor from seeking additional remedies from customers and end users of a licensed technology. These types of clauses can provide even broader protection than the first sale or exhaustion doctrine because they prevent the licensor from asserting their IP against third parties even when the licensee’s technology satisfies only one particular element of a patent claim or one particular element of a cause of action.

Liability for Past Damages; Releases

Although an explicit license grant and a no-assertion clause can protect licensees and their business partners, customers, and suppliers on a going-forward basis, they do not provide protection from potential liability for past damages. In Cascade Computer Innovation LLC v Samsung Electronics Co. (ND Ill 2014) 70 F Supp 3d 863, the plaintiff asserted that a number of Android mobile device manufacturers infringed the patent-in-suit because they practiced the claimed method by using the Dalvik virtual machine found in the Android operating system. Google entered into a license with Cascade, and Samsung and HTC, two of the Android mobile device manufacturers, moved for summary judgment of no liability as a result of the Google license. The court determined that Samsung and HTC were free from forward-looking liability but not past damages, stating (70 F Supp 3d at 869):

Patent exhaustion is a rule that looks forward; it results from the patent holder’s sale of a patented item or a product embodying a patented method, or a grant of a license permitting another to use the patented item or method. A release, by contrast, looks backward and takes an alleged infringer off the hook for something it has already done or is alleged to have done.

In Cascade, the release and covenant not to sue were specifically limited to Google and its affiliated entities. As the court pointed out, “a release given to one tortfeasor does not release others and does not, unlike a sale or a license, surrender the releasor’s rights.” 70 F Supp 3d at 870.

As a result, to ensure that a licensee’s business partners, suppliers, and customers are truly free from all potential liability for all time arising from their use of or participation in supplying the accused technology, drafters should include a carefully worded release. In particular, a release should specify precisely the actors who are released from all liability. These actors can be identified by name, categories (e.g., software vendors or retailers), or by a defined term such as “Released Entities” that includes categories and specific roles (e.g., supplying the user interface technology or supplying the database technology) filled by these business partners. As discussed above, a licensor will want to take great care to limit the release specifically to the licensed technology to avoid inadvertently releasing all the downstream users, business partners, and suppliers from all potential liabilities arising from technology unrelated to what is being licensed.

Parties who have decided to designate California law as the choice of law for their license agreement should also be aware of the so-called section 1542 waiver. Civil Code §1542 states:

A general release does not extend to claims which the creditor does not know or suspect to exist in his or her favor at the time of executing the release, which if known by him or her must have materially affected his or her settlement with the debtor.

Licensees and beneficiaries of a license benefit from an explicit waiver of the rights provided under §1542 because the waiver forecloses the possibility of any future litigation should the licensor realize at a later point that it has another claim. However, such a waiver may conflict with a carefully crafted release that is limited to activities pertaining to the supplying or use of the licensed technology. Thus, parties should ensure that they reach agreement on whether a section 1542 waiver should be included in a license before finalizing and executing a license agreement.

Limitations on Scope of the License

Imposing limits on the scope of a license can be a difficult and uncertain challenge. Traditionally, the U.S. Supreme Court had held that “the patentee may grant a license ‘upon any condition the performance of which is reasonably within the reward which the patentee by the grant of the patent is entitled to secure.’” General Talking Pictures Corp. v Western Elec. Co., (1938) 305 US 124, 127 (citation omitted). In a companion case, the American Transformer Company was a licensee who received a license to manufacture and use patented amplifiers for “radio amateur reception, radio experimental reception, and home broadcast reception. It had no right to sell the amplifiers for use in theaters as a part of talking picture equipment.” General Talking Pictures Corp. v Western Elec. Co. (1938) 304 US 175, 180. Despite explicit limits in the license, the licensee made and sold the patented amplifiers to General Talking Pictures to be used in the motion picture industry. The Supreme Court held that the manufacture and sale of patented amplifiers to General Talking Pictures was explicitly outside the scope of the license grant given to the licensee. Therefore, the sale of the amplifiers to General Talking Pictures was not licensed and constituted an act of infringement.

Seventy years later, the Supreme Court revisited patent licenses in Quanta Computer, Inc. v LG Electronics, Inc. (2008) 553 US 617, and was careful to distinguish between the license in that case with the license in General Talking Pictures. In Quanta, the Supreme Court specifically noted that (553 US at 636):

[n]othing in the License Agreement restricts Intel’s right to sell its microprocessors and chipsets to purchasers who intend to combine them with non-Intel parts. It broadly permits Intel to ‘make, use, [or] sell’ products free of LGE’s patent claims.... Hence, Intel’s authority to sell its products embodying the LGE Patents was not conditioned on the notice or on Quanta’s decision to abide by LGE’s directions in that notice.

In parsing the LG-Intel license, the Supreme Court determined that LG’s patent rights were exhausted because Intel’s microchip components substantially embodied the patents-in-suit and that LG could not control Quanta’s subsequent combination of licensed Intel components with non-licensed Quanta components in a Quanta-branded computer. 553 US at 638.

The Quanta decision was viewed by many as a sea change, indicating a curtailing of the licensor’s ability to impose conditions or limits on the scope of the license recognized in General Talking Pictures. The recent Cascades decision reaffirms this general trend. In Cascades, the licensor attempted to restrict the license to only Google and Motorola. To effectuate that intent, the license specifically stated that Google received a license to practice the patent in any “Google Product.” Cascades Computer Innovation LLC v Samsung Electronics. Co., Ltd. (ND Ill 2014) 70 F Supp 3d 863, 865. In turn, “Google Product” was defined as any “product[] of Google, Motorola and/or Google Affiliates, including all Motorola and Nexus devices, but ... exclud[ing] mobile devices manufactured by third parties and running the Android OS except any Nexus-branded devices.” 70 F Supp 3d at 865 (emphasis added). The Cascade Court reasoned that (70 F Supp 3d at 868):

Google was authorized to convey to others, including Samsung and HTC, products—including the Android operating system—that practiced Cascades’s patents. As a result, Cascades could no longer assert patent rights with respect to those products. As was the case in Quanta, use of the restriction in the Cascades/Google license agreement to limit how those who thereafter acquired the Android operating system from Google could use it would in effect allow Cascades to circumvent the patent exhaustion doctrine and reap multiple gains from a single sale.

Despite the explicit clause in the Cascades-Google license excluding third parties running the Android operating system, the court determined that the Android operating system would fall under the definition of “Google Product.” As a result, Google was free to distribute the Android operating system to third party mobile device manufacturers, who in turn were free to use Android without any liability. Under the guise of patent exhaustion, the court re-wrote the license to remove an explicit limitation that the parties had agreed on.

The Federal Circuit did recently uphold some scope-limiting language in a patent license, and in so doing, placed limits on the breadth of patent exhaustion. In Helferich Patent Licensing, LLC v New York Times Co. (Fed Cir 2015) 778 F3d 1293, the asserted patents described two sets of related wireless communication technologies. The first set of claims was directed towards wireless handsets that received and requested certain content (handset claims). A second set of claims was directed toward methods and systems for storing and updating information of various types and sending it to handsets (content claims). Helferich had previously licensed all the handset companies under the handset claims and explicitly carved out any rights to the content claims. Helferich then brought suit against a number of media companies, asserting they infringed the content claims by providing content to users on their mobile handsets. The trial court found that Helferich’s patent rights were exhausted with respect to the media companies because it had previously licensed its portfolio to all handset manufacturers, and the media companies presumably infringed by sending content to mobile handsets.

However, the Federal Circuit reversed, noting that the district court failed to focus on the specific language of the claims. In particular, the Federal Circuit pointed out two deficiencies in the defendants’ arguments. First, the media companies did not (and could not) allege that the licensed handset companies actually practiced any of the content claims. Second, defendants did not argue that the content claims required the use of licensed handsets that contain the inventive features of the handset claims. The fact that content claims may in some way contemplate use of the handset claims is not sufficient to establish that the content claims have been exhausted by a license of the handset claims.

The Federal Circuit reasoned that Helferich’s licenses “reflect painstaking efforts to distinguish the conduct of handset makers and possessors from the conduct of others, such as content providers, and to distinguish claims practiced by the former from claims practiced by the latter.” 778 F3d at 1297. In the portfolio license with the handset makers, the “Licensed Fields” were defined as “Mobile Wireless Communication Devices” made, used, or sold by the manufacturer licensee. 778 F3d at 1297. Notably, the definition of “Licensed Fields” disclaimed any grant of rights to content providers and reserved Helferich’s right to enforce against them.

In view of Helferich, when a patent has multiple related but separately patentable inventions, a patent holder is free to license those inventions separately. To protect this interest, patent holders should take care to structure their license agreements along the lines that Helferich did, clearly demarcating and naming the different inventions and designating the specific claims that belong to each invention. Additionally, the license should state any particular claims or inventions that are not licensed to a particular licensee. Patent prosecutors can also take care to separate out groups of claims to distinctly point out that there are multiple related but separately patentable inventions.

Decisions like Quanta and Cascades demonstrate that any attempt to exclude specific customers or downstream users of a licensed technology will likely fail. However, other types of limitations, such as the field-of-use restrictions upheld in General Talking Pictures and the separate licensing for complementary but non-identical inventions described in Helferich, remain viable options, but they will be scrutinized carefully by the courts given recent trends in the law.


No other clause in a license agreement has a more obvious impact on a litigator’s job than a dispute resolution clause. When disputes arise, the dispute resolution clause can have profound implications on a party’s ability to vindicate its rights under a license. Similarly, the indemnity clause can shift risk to better protect either the licensee or licensor from third party claims and other potential harm. Despite these implications, these clauses are often overlooked during drafting and negotiation because they may seem less important next to core business terms such as the scope of licensed rights and royalty rates. In drafting, licensors and licensees should take time to think about the types of disputes that are likely to arise under the license, what forum and law are most appropriate to resolve those disputes, and which party should bear the risk of defending the licensed activities against third-party claims.

Choice of Forum: Litigation Versus Alternative Dispute Resolution

Forum Selection Clauses

Parties to patent licenses may resolve their disputes through court litigation or arbitration. A forum selection clause is enforceable if it arises from a “freely negotiated” contract and is not “unreasonable and unjust.” Burger King Corp. v Rudzewicz (1985) 471 US 462, 472 n14. See also Atlantic Marine Constr. Co., Inc. v District Court (2013) ___ US ___, 134 S Ct 568, 582 (“forum selection clauses should control except in unusual cases”); Monsanto Co. v McFarling (Fed Cir 2002) 302 F3d 1291, 1296.

When a litigation forum is chosen, the court still must determine that it has subject matter jurisdiction over the dispute; parties cannot create subject matter jurisdiction by agreement. A forum selection clause choosing a particular court essentially just forecloses a party’s objection to the selected court’s personal jurisdiction over that party.


Parties may also choose to arbitrate disputes, even those under a patent license. The Patent Act (35 USC §§100-376) explicitly allows parties to resolve disputes relating to patent validity or infringement through arbitration, which “shall be final and binding between the parties to the arbitration but shall have no force or effect on any other person.” 35 USC §294(c). In such an arbitration, a party may raise defenses provided for by 35 USC §282, such as invalidity and unenforceability, and the arbitrator must consider those defenses. 35 USC §294(b). Courts will thus enforce agreements to arbitrate patent licenses. Deprenyl Animal Health, Inc. v Univ. of Toronto Innovations Found. (Fed Cir 2002) 297 F3d 1343, 1358 (staying a declaratory judgment action seeking invalidity and non-infringement of licensed patents pending outcome of Canadian arbitration of license agreement); Conteyor Multibag Sys. N. V. v. Bradford Co. (WD Mich, Aug. 10, 2006, No. 1:05-CV-613) 2006 US Dist Lexis 55898 (compelling arbitration when licensor had brought both a district court action for infringement of a licensed patent and then later an arbitration of license agreement, and licensee had asserted counterclaims in district court action for invalidity and unenforceability of patent).


Although parties to a contract can select a particular court to hear their disputes, they cannot select a particular judge of that court. Nevertheless, one of the benefits of choosing a public court as the forum is that there are resources, such as Westlaw, PACER, and Docket Navigator, to research how the judge assigned to a particular case has managed or decided similar disputes in the past. These resources can prove invaluable for litigation counsel when crafting and advising clients on litigation strategy and budgeting.

On the other hand, although some arbitration awards are published, many are not. The lack of published decisions results from the choice of many parties to conduct arbitration confidentially—a choice not available for commercial disputes in public courts. (Parties may of course seek to seal documents filed with courts, and even to seal hearings, but the ultimate decision to seal lies with the court, not with the parties.) It may therefore be impossible to find any published arbitral award issued by a particular arbitrator. Litigation counsel may be able to gather anecdotal evidence from colleagues who have experience with a particular arbitrator, but this type of information is obviously not of the same predictive quality as published opinions.

Nevertheless, like other aspects of the dispute resolution clause, the parties are free to specify the qualifications of any arbitrator who may be appointed, and most of the major bodies that administer commercial arbitrations will do their best to find someone who meets those qualifications. See, e.g., American Arbitration Ass’n Commercial Arbitration Rules and Mediation Procedures (AAA Rules), R-13; International Arbitration Rules of the International Chamber of Commerce (ICC Rules), Art. 12. Obviously, the more requirements there are, the harder it will be to find someone who meets them all. However, the administering bodies will not let the fact that no one meets the specified requirements prevent them from appointing the best arbitrator they can find to decide the dispute.

It is worth considering the most important qualities that litigation counsel and the client will want in the neutral, who may be deciding whether the client will get paid—or have to pay—substantial royalties. For example, if the licensed patents are U.S. patents, it may make sense to require that the arbitrator be a retired U.S. district judge who has presided over patent trials. The client may not want a novice or foreigner deciding nuanced issues of claim construction or invalidity under U.S. patent law.

Procedural Formality

One of the most striking differences between litigation and arbitration is procedural formality. Civil litigation in federal courts is governed by a host of procedural rules, including the Federal Rules of Civil Procedure, the Federal Rules of Evidence, and local rules of the federal district court. Many district courts also have local patent rules. Similarly, state courts in California are governed by the Code of Civil Procedure, Evidence Code, California Rules of Court, and their own local rules. When parties choose litigation, these procedural rules will provide litigation counsel with some measure of predictability regarding issues such as the mechanisms, timing, scope, and sequence of discovery. This predictability, in turn, will allow litigation counsel to plan for what evidence he or she will need and be able to obtain through discovery to prove his or her client’s case and how to budget for it.

In contrast, rules of procedure in arbitration are often left to the broad discretion of the arbitrator(s). For example, parties to cross-border licenses often choose the ICC Rules to govern their disputes. Under Article 22(2) of the ICC Rules,

[i]n order to ensure effective case management, the arbitral tribunal, after consulting the parties, may adopt such procedural measures as it considers appropriate, provided that they are not contrary to any agreement of the parties.

Similarly, the ICC Rules’ guidance on discovery and evidence is stated in Article 25(1) as follows: “The arbitral tribunal shall proceed within as short a time as possible to establish the facts of the case by all appropriate means.” Under the AAA Rules, “[c]onformity to legal rules of evidence shall not be necessary.” AAA Rules, R-34(a). Even the seemingly straightforward question of the language of the arbitration cannot be taken for granted if not agreed to by the parties. See ICC Rules, Art. 20.

When choosing arbitration, then, parties should specify the arbitral procedure with as much detail as possible. For example, if a cross-border patent license dispute arises, in the absence of an agreement on discovery procedure an American arbitrator may be willing to allow broad, American-style discovery, while a European arbitrator may not. In the latter case, litigation counsel may find himself or herself unable to prove an essential element of a claim or defense. On the other hand, a party may wish to avoid certain discovery mechanisms altogether (e.g., allowing production of documents but not depositions), whether to minimize costs or inconvenience to its employees or simply to make it more difficult for the other side to prove its case. Because reasonable agreements on procedure are likely to be followed by an arbitrator (see ICC Rules, Art 22(2)), it behooves the parties to specify their desired procedure to the extent that they can.

How to decide which procedure should govern? A little time spent thinking about the most likely disputes to arise under the license and what the client will need to prove its claims or defenses regarding that dispute should help. For example, patent license disputes often concern whether royalties are owed on particular products, which issue typically turns on the question of whether those products would, but for the license, infringe the licensed patents. What kind of evidence will the client need to prove or disprove that issue? If counsel does not know, counsel should ask a patent litigation colleague.


In federal and state courts, litigants may appeal from an adverse judgment in a trial court as a matter of right. In addition to the rules governing trial court procedures, there are appellate rules governing how trial counsel must preserve issues for appeal and how to make an appeal. The appellate court’s decision, like the trial court’s, will be guided by the law. An adverse trial court judgment is thus not necessarily the last word in a dispute.

In contrast, it is difficult for litigation counsel to overturn an unfavorable arbitral award. Under the Federal Arbitration Act (FAA) (9 USC §§1-16), which incorporates the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (see 9 USC §§201-208), an arbitral award may be challenged by the losing party on very limited grounds. For a domestic arbitration subject to the FAA, a court may vacate an arbitral award when it was obtained through “corruption, fraud, or undue means,” when there was “evident partiality or corruption in the arbitrators,” when the “arbitrators were guilty of misconduct,” or when the “arbitrators exceeded their powers.” 9 USC §10. The bases for which a court may refuse to enforce a foreign arbitral award are similarly narrow. See 9 USC §207; Convention on the Recognition and Enforcement of Foreign Arbitral Awards art 5, entered into force for U.S. Dec. 29, 1970, 21 UST 2517, 330 UNTS 38.

Appealability (or the lack thereof) may be precisely why parties to license agreements choose litigation over arbitration or vice versa. Licensors seeking royalties may be more interested in a decision based on the equities of a particular case. They may prefer a single arbitrator to a three-arbitrator panel in the belief that a panel is less likely to reach a decision based on equities. They may not want the arbitrator to issue a reasoned award lest it provide grounds for a court to vacate the award later. Licensees may want the opposite of these factors. It is important for the client to understand which position it is likely to be in if a dispute arises and what it will want then. Those considerations should weigh heavily in the choice of litigation versus arbitration.

Cost and Time

The conventional wisdom is that arbitration is faster and cheaper than litigation. This may or may not be true in any given case, especially a case involving patents on complex technology. See, e.g., Deprenyl Animal Health, Inc. v Univ. of Toronto Innovations Found. (Fed Cir 2002) 297 F3d 1343, 1358 (directing stay of declaratory judgment action in favor of Canadian arbitration proceeding concerning patent license agreement 2 years after arbitration commenced). See also Sanofi-Aventis Deutschland GmbH v Genentech, Inc. (Fed Cir 2013) 716 F3d 586, discussed below (affirming denial of motion to enjoin arbitration of patent license agreement more than 4 1/2 years after arbitration commenced). Of course, litigation—especially patent litigation—can drag on in the trial and appellate courts for years and cost the client dearly.

For the reasons discussed above and others, the client should understand that selection of arbitration over litigation will not necessarily result in a faster or less expensive dispute resolution. In fact, when arbitration is selected, litigation counsel, who already faces a daunting task in predicting how litigation will unfold, is even less able to predict the timelines and expenses that will be involved. The client should understand this difficulty before the dispute resolution clause is finalized.

Choice of Law

As a general rule, a contractual choice-of-law clause will be enforced if it has a substantial relationship to the parties or the contract, e.g., where performance is to take place or where one of the parties is incorporated. See Nedlloyd Lines B.V. v Superior Court (1992) 3 C4th 459, 462. See also CC §1646.5 (providing for enforcement of a California choice-of-law clause in transactions of at least $250,000). Parties to patent license agreements often simply choose the law of the home jurisdiction of one or more of them as a default, as they would for any other contract, without giving much thought to how this choice might actually affect disputes down the road. The parties’ choice of law can have substantial consequences, as the case below demonstrates.

In Sanofi-Aventis Deutschland GmbH v Genentech, Inc. (Fed Cir 2013) 716 F3d 586, German law governed the patent license at issue, which called for arbitration of disputes in accordance with the ICC Rules. 716 F3d at 589. The license required the licensee to pay the licensor royalties on its sale of “materials . . . the manufacture, use or sale of which would, in the absence of this Agreement, infringe one or more unexpired issued claims of the” licensed patents. 716 F3d at 589 (emphasis added). The licensee refused to pay royalties on products that the licensor claimed infringed the licensed U.S. patents, and the licensor demanded arbitration. The licensee then terminated the license and brought an action in the federal district court for a declaratory judgment that the licensed patents were invalid and that the accused products did not infringe them. The foreign arbitration and U.S. litigation proceeded in parallel, until the court eventually determined that the accused products did not infringe the licensed patents under U.S. law and the Federal Circuit affirmed that finding. The licensee then moved the federal district court to enjoin the licensor from continuing with the foreign arbitration, arguing that the judgment of non-infringement disposed of all issues in the arbitration. 716 F3d at 590. Meanwhile, the arbitrator decided that, under German law, the licensee did infringe the U.S. licensed patents. In affirming the federal district court’s denial of an injunction against the arbitration, the Federal Circuit concluded that “the U.S. judgment of non-infringement is not dispositive as to breach of the Agreement.” 716 F3d at 593. It found that “the meaning of infringement under the Agreement and the meaning of infringement under U.S. law are not functionally the same.” 716 F3d at 593.

Licensing counsel and the client should consider what will be the most likely dispute to arise under the license. In many cases, it will be a question of infringement. Counsel should consider whether he or she knows what principles would apply to decide that issue under the proposed jurisdiction’s law. If the client does not like those principles, counsel should propose different law. For example, the client may want questions of infringement of U.S. patents decided under U.S. law, instead of a foreign country’s laws. If so, counsel should ensure that the license reflects that decision.


In IP license negotiations, especially those between a supplier and a customer in a manufacturing chain, the scope of the indemnity clause will be a subject of much debate. The customer will seek to broaden the indemnity clause to cover all possible assertions of intellectual property against the licensed technology and related losses, regardless of whether the licensed technology forms only one component or one portion of an overall product. The supplier will often push in the opposite direction to limit the indemnity clause to cover only third party claims that are directed at the licensed component and nothing more. To clarify this limitation, suppliers may ask for an indemnity clause that explicitly carves out indemnity coverage when the licensed component is combined with other non-licensed components to create a larger system or device.

The factual situation in the Quanta case discussed above presents one such example. In Quanta, the licensed component was a microchip inside a laptop. In this situation, the laptop retailer will want a broad indemnity clause from the chip supplier so that indemnity would still apply even if the intellectual property that is asserted against the laptop arguably implicates a functionality provided by the chip. The chip supplier will seek to avoid providing indemnity for the entire laptop and try to limit the indemnity to assertions only against the chip. The breadth of this clause will have a big impact down the road should a third party file a claim against the laptop.

In addition to the breadth of the indemnity clause and covered losses, important consideration should be paid to terms addressing which party has the right to select counsel, the power of each party to direct or participate in litigation, and who has authority to enter into a settlement and under what circumstances. For example, rather than using the boilerplate term “hold harmless,” the effect of which can vary under state common law, drafters should consider explicitly stating the scope of loss coverage. Similarly, because indemnification clauses have a direct impact on how litigation defense is conducted and who gets to control the litigation, drafters should take care to set out the procedural and substantive requirements for indemnification and the exceptions. When the licensed technology is only a limited piece of the ultimate product, licensors should make sure the indemnification provisions are drafted to limit their indemnification obligations accordingly.


Exclusive Licenses and Exclusive Dealing Arrangements

Intellectual property license agreements can be exclusive as to either the license grant itself (an exclusive license) or with respect to requirements imposed on the licensee. When the licensor enters into an exclusive license, only one licensee can practice the licensed technology. Exclusive licenses are generally permitted under the antitrust laws, although there are several exceptions when the license effectively blocks competition in a particular market. See U.S. Dep’t of Justice and Federal Trade Comm’n, Antitrust Guidelines for the Licensing of Intellectual Property (1995) (IP Guidelines), §4.1.2. For example, there may be antitrust concerns when an exclusive license is between horizontal competitors and it eliminates sufficient market competition. IP Guidelines, §4.1.2.

The licensor may also use a contract to impose exclusive dealing requirements on the licensee—for example, prohibiting the licensee from using or distributing competing technologies for a set duration, or using contingent incentives in an effort to obtain a similar effect. Antitrust law recognizes de facto exclusivity based on the real world impact of the agreement; therefore, a contract does not need to include an explicit exclusive dealing requirement to create harm to competition. R. J. Reynolds Tobacco Co. v Philip Morris Inc. (MD NC 2002) 199 F Supp 2d 362, 387. De facto exclusive dealing claims have been alleged against contractual provisions such as conditional rebates, preferential pricing requirements, and restrictions with respect to competitors. Along with explicit exclusive dealing requirements, these type of clauses are permitted in a range of circumstances and provide licensors with a tool to help promote the adoption and use of a particular technology, as well as incentivize development on top of the licensed technology. See IP Guidelines, §5.4.

Exclusive dealing typically deserves more careful consideration at the drafting stage. As discussed below, based on the competitive landscape and the licensor’s position in it, the licensor can often use a number of different exclusive dealing provisions in structuring an IP license. In contrast, the antitrust and regulatory issues around exclusive licensing more commonly arise not in drafting, but with respect to pre-licensing conduct, mergers, and refusals to deal, all of which are outside the scope of this article.

Defining the Relevant Market

For exclusive dealing, antitrust law is focused on conduct that either harms consumers in the relevant market or stifles market competition for future products and innovations. See U.S. Dep’t of Justice, Antitrust Enforcement in High-Technology Industries: Protecting Innovation and Competition, Remarks of Fiona Scott-Morton prepared for 2012 NYSBA Annual Antitrust Forum, Dec. 7, 2012, available at When considering exclusive dealing provisions, defining the relevant market is the starting place for evaluating antitrust risk. In addition to product and service markets, the relevant antitrust market for patents and other intellectual property can be that for a particular technology, or even innovation around that technology. See IP Guidelines, §3.2.3. The market for a product or technology is defined relative to the surrounding market based on “the cross-elasticity of demand for,” or “reasonable interchangeability of,” related products or technologies. See Coalition for ICANN Transparency, Inc. v Verisign, Inc. (9th Cir 2010) 611 F3d 495, 507 (citations and internal quotation marks omitted). Cross-elasticity of demand is an economics concept that measures how and the extent to which demand for other products or services is affected by an increase or decrease in price, e.g., does steep discounting of one product or service lead consumers to switch to that product or service as a substitute for another product? See ABA Section of Antitrust Law, Antitrust Law Developments (6th ed 2007) (ABA Antitrust Developments), at 562. When an increase in the licensor’s price drives consumer demand to other products, those other products are typically considered part of the relevant market under a cross-elasticity analysis. In addition, the market definition assessment of reasonable interchangeability is non-quantitative. Courts will consider things such as public recognition of a submarket, customer dynamics, and particular aspects of the product or service that may not be captured in a purely quantitative analysis. ABA Antitrust Developments, at 563.

In evaluating the market boundary for the product or service to be licensed, it may be appropriate to consider a broad definition of potential competitors or substitutes. The market definition analysis in U.S. v Oracle Corp. (ND Cal 2004) 331 F Supp 2d 1098, a government suit challenging a merger, shows the complexity and importance of looking beyond immediate competitors. In Oracle, the federal government and several states sought to block a merger between Oracle and PeopleSoft on the grounds that it was likely to substantially reduce competition in a specific market in violation of §7 of the Clayton Antitrust Act (15 USC §18). The plaintiffs alleged that the relevant market was that for two types of enterprise resource planning software: human relations management (HRM) and financial management systems (FMS). They also tried to limit the market to “high function” HRM and FMS software. Following a trial, the court sided with Oracle and rejected the government’s market definition on multiple grounds. Among other things, the court held that outsourcing the function that the Oracle software was intended to provide was part of the relevant market. Substitutes therefore went far beyond directly comparable systems to include alternative means for consumers to accomplish the same objective.

Market Power and Foreclosing Competition

An exclusive dealing agreement, or series of agreements, violates §1 of the Sherman Antitrust Act (15 USC §1) only when the “effect is to ‘foreclose competition in a substantial share’” of the relevant market. Allied Orthopedic Appliances Inc. v Tyco Health Care Group LP (9th Cir 2010) 592 F3d 991, 996 (citations omitted). This determination is made based on the rule of reason. Section 3 of the Clayton Antitrust Act (15 USC §14) does not require an actual foreclosure of competition; instead, exclusive arrangements can be unlawful when the “probable effect” is to foreclose substantial competition. 592 F3d at 1003 n2. The analysis for both types of claims incorporates the economic and competition benefits of the arrangement. For example, an exclusive dealing requirement may be necessary for the licensor to ensure that it can recoup high up-front costs. Exclusive arrangements may also be relevant to §2 of the Sherman Antitrust Act (15 USC §2) when they are used to maintain monopoly power. See U.S. v Microsoft Corp. (DC Cir 2001) 253 F3d 34.

To foreclose competition, the licensor must have market power. See Murrow Furniture Galleries, Inc. v Thomasville Furniture Indus., Inc. (4th Cir 1989) 889 F2d 524, 528. Market power is the “the ability to raise prices above the levels that would be charged in a competitive market.” R. J. Reynolds Tobacco Co. v Philip Morris Inc. (MD NC 2002) 199 F Supp 2d 362, 381 (citations omitted). Although market power is based on a variety of factors specific to the particular circumstances and is often dependent on circumstantial evidence, market share is usually one of the threshold considerations in evaluating market power. As a general rule, a licensor market share of around 70-75 percent is considered necessary to support a market power finding. See 199 F Supp 2d at 394 (collecting cases). Thus, smaller licensors are typically not limited by antitrust concerns in using exclusive dealing provisions.

In R.J. Reynolds, the plaintiffs alleged that Philip Morris’s retail merchandising program violated §§1 and 2 of the Sherman Antitrust Act (15 USC §§1-2), along with other state laws. The Philip Morris retail distribution agreements at issue had provisions intended to increase sales. Under the distribution program, participating retailers could obtain significant discounts based on their participation level, including retailer display, advertising, and promotional space. R.J. Reynolds, 199 F Supp 2d at 370. In places where Philip Morris had at least 55 percent of local market share, it required a 90 percent share of product space. At the highest level of participation, there was typically room for only one competitor at the point of sale. 199 F Supp 2d at 370. In all cases, however, retailers could sell competitive products. Following implementation of this program, Philip Morris’ market share increased from around 49.6 percent to 51 percent. 199 F Supp 2d at 372.

In granting summary judgment in favor of defendant Philip Morris, the court found there was “no evidence in the record to suggest that Marlboro smokers will not switch to a lower-priced brand if Marlboro is not priced competitively.” 199 F Supp 2d at 373. Thus, despite all of the various aspects of the marketing program, there was no direct evidence that Philip Morris had sufficient market power to impose ultra-competitive prices. 199 F Supp 2d at 381. In addition, R.J. Reynolds could not show sufficient circumstantial evidence of market power because the Philip Morris marketing programs were implemented in a market with declining demand and excess capacity; existing competitors were therefore a valid check against Philip Morris. 199 F Supp 2d at 382.

Even when the licensor has market power, courts typically require that close to half of the relevant market has been foreclosed, although high barriers to entry can lower the requisite foreclosure. See U.S. v Microsoft Corp. (DC Cir 2001) 253 F3d 34, 70 (noting that “roughly 40% or 50% share [foreclosure is] usually required in order to establish a §1 violation”). For example, in Philip Morris, the court noted that even assuming the Philip Morris agreements foreclosed competition in 34 percent of the relevant market, this was not a “substantial share of the relevant market” under the circumstances. 199 F Supp 2d at 388 (citations omitted). Therefore, in drafting, licenses that will cover only a fraction of the relevant market can typically include provisions directed at raising barriers for competitors.

Term of Licensing Agreement

Term language in exclusive licensing agreements, including the duration of the agreement and the ability of either the licensor or the licensee to terminate the agreement, is one of the most significant considerations for avoiding antitrust harm. Agreements with short durations present a minimal threat to foreclosing competition. See ZF Meritor, LLC v Eaton Corp. (3d Cir 2012) 696 F3d 254, 286 (citing Christofferson Dairy, Inc. v MMM Sales, Inc. (9th Cir 1988) 849 F2d 1168, 1173). For example, the court in Barry Wright Corp. v ITT Grinnell Corp. (1st Cir 1983) 724 F2d 227, 237, found 2-year contracts to be reasonable. See also Western Parcel Express v UPS of America, Inc. (ND Cal 1998) 65 F Supp 2d 1052, 1064 (referring to contracts ranging from 30 days to 3 years as having “relatively short duration”). For agreements with longer terms, easy terminability of the arrangement, e.g., the ability to terminate for convenience on limited notice after an initial period, can negate the effect of an exclusive dealing arrangement. See Omega Envt’l, Inc. v Gilbarco, Inc. (9th Cir 1997) 127 F3d 1157, 1162. In considering the term and terminability of an exclusive dealing contract, business needs for a particular agreement, e.g., to recoup high up-front investment costs for the licensor, may also be used to justify the arrangement as pro-competitive.

In R.J. Reynolds, even though Philip Morris used a number of provisions to increase sales over its competitors, there was no antitrust harm in light of the strong market competition left in place after the agreements. See R.J. Reynolds Tobacco Co. v Philip Morris Inc. (MD NC 2002) 199 F Supp 2d 362. In contrast, in ZF Meritor, Eaton was liable under the Sherman Antitrust Act when its distribution agreements left only 15 percent of the remaining market open. ZF Meritor, LLC v Eaton Corp. (3d Cir 2012) 696 F3d 254, 286. In ZF Meritor, antitrust defendant Eaton Corp. distributed heavy-duty truck transmissions to truck buyers through original equipment manufacturers (OEMs). After a period of declining demand, Eaton entered into long-term supply agreements with each of the four OEMs; each agreement had a term of at least 5 years. Among other provisions, each OEM agreement provided conditional rebates for the OEM if it met certain purchase requirements; the highest purchase requirement was 97.5 percent. The OEM agreements also required that Eaton’s products have preferential pricing over competing products, and two of the agreements required suppliers to remove competing products from the product data book for sales. 696 F3d at 265.

Following a jury trial almost a decade after the relevant agreements had been executed, the court concluded that there was sufficient evidence for the jury’s findings that Eaton’s conduct violated §§1 and 2 of the Sherman Antitrust Act (15 USC §§1-2), and §3 of the Clayton Antitrust Act (15 USC §14). The court awarded no damages, however, after rejecting the plaintiff’s expert damages testimony and denying a request to amend the report. 696 F3d at 267. On appeal, the Third Circuit upheld the trial court’s decision that the jury verdict was sufficiently supported. In affirming the jury decision, the court noted that, given Eaton’s size, losing it as a supplier “was not an option” for the OEMs, and thus they were forced to agree to things they might otherwise have avoided. 696 F3d at 278. For example, Eaton could require the OEMs to remove competing products from the data book listing, even though this was uncommon in the industry and detrimental to OEM customers according to the evidence at trial. 696 F3d at 277. In addition, it was “unprecedented for a supplier” to lock up over 85 percent of the market with contracts of at least 5 years. 696 F3d at 287. In combination with the other factors, this left no room for meaningful competition, and so this was “the rare case in which exclusive dealing would pose a threat to competition.” 696 F3d at 285 (citation omitted).

Before including exclusive dealing provisions, however, the licensor should evaluate the extent to which the license or series of licenses will leave open the potential for consumers to obtain competing products or use competing technologies. Provided that competitors still have other distribution options to reach the ultimate consumers, the antitrust risk is minimal.

Licensing Through Distribution Agreements

As a final note, both R.J. Reynolds and ZF Meritor involved distribution agreements. When licensing intellectual property through distribution agreements, the licensor can often use contracts to restrict distribution channels for competitors. Before including exclusive dealing provisions, however, the licensor should evaluate the extent to which the license or series of licenses will leave open the potential for consumers to obtain competing products or use competing technologies. Provided that competitors still have other distribution options to reach the ultimate consumers, the antitrust risk is minimal. See Omega Envt’l, Inc. v Gilbarco, Inc. (9th Cir 1997) 127 F3d 1157, 1164. In ZF Meritor, for example, the defendant entered into agreements with all OEMs in the market, and sales in that market were not done directly to consumers. See ZF Meritor, 696 F3d at 265. But in R.J. Reynolds, where the court ruled for the defendant on summary judgment, the exclusive dealing provisions did not prevent consumers from buying competing products, even though Philip Morris tried to make it more difficult to compete at the distributor level. See R.J. Reynolds, 199 F Supp 2d 362, 370.


Drafting an intellectual property license requires careful consideration of a number of factors, including the scope of the license, the upstream and downstream implications, the rules governing dispute resolution, and the allocation of risk of loss arising from third party claims against the licensed technology. Although these clauses have an obvious impact on the efficacy of a license, other terms setting forth market incentives, rebates, or pricing requirements can have an equally important impact. These market-based terms can be used to increase distribution of the licensed technology or limit the market for competitors. Although there are rare circumstances when exclusive dealing provisions violate antitrust laws, these types of provisions can be used by most licensors, particularly newer and smaller licensors.

Firm Highlights


Takeaways From the Proposed PREVAIL Act

Two proposed bills recently introduced in Congress have the potential to greatly impact the current patent litigation landscape. The bills are titled the Promoting and Respecting Economically Vital American Innovation Leadership (PREVAIL) Act and the Patent...

Read More

AI and Trade Secrets: A Complicated Friendship

Excitement has been growing for decades around the development, training, and use of generative AI, but this past year the excitement escalated into a frenzy. Everyone is considering how AI impacts their business. This article...

Read More

Farella Lawyers Recognized as 2023 IP STARS by Managing Intellectual Property

Farella Braun + Martel intellectual litigation partners Daniel Callaway , James Day , Jeffrey Fisher , Winston Liaw , and Eugene Mar were recognized by Managing Intellectual Property in the 2023 edition of IP...

Read More

It Wasn’t Me, It Was the AI: Intellectual Property and Data Privacy Concerns With Nonprofits’ Use of Artificial Intelligence Systems

In today's rapidly changing technological landscape, artificial intelligence (AI) is making headlines and being discussed constantly. To be sure, AI provides a powerful tool to nonprofits in creating content and exploiting for countless cost-effective...

Read More

USAA Wins Patent Agreement With Discover, Vows To Keep Going After Banks

Eugene Mar spoke to American Banker for the article, "USAA Wins Patent Agreement With Discover, Vows To Keep Going After Banks." Link (subscription required): Eugene said this becomes a business decision for a...

Read More

Farella Braun + Martel Attorneys Named to 2023 Northern California Super Lawyers and Rising Stars

Thirty-eight Farella Braun + Martel lawyers were named to the Super Lawyers and Rising Stars lists of top attorneys in Northern California for 2023. 2023 Farella Northern California Super Lawyers: Carly Alameda – Business...

Read More

What Patent Bills Would Mean for Infringement Litigation

Two bills recently introduced in Congress could significantly affect the current patent litigation landscape. The bipartisan bills are titled the Patent Eligibility Restoration Act of 2023 and the Promoting and Respecting Economically Vital American...

Read More

Farella Lawyers Recognized in The Best Lawyers in America® 2024 Edition

Read More

Fair Use Question Goes to Trial in AI Copyright Lawsuit – Thomson Reuters v. Ross Intelligence

On September 25, 2023, a United States Circuit Judge determined that fact questions surrounding issues of fair use and tortious interference required a jury to decide media conglomerate Thomson Reuters’s lawsuit against Ross Intelligence...

Read More

Farella Braun + Martel Earns 2024 Best Law Firms® Rankings

Read More