Difference between revisions of "Patent License Drafter's Notes"

Drafter’s Notes

See also Model Patent License which is superseded by CC Public Patent License

Purpose:

These Drafter’s Notes (“Notes”) capture many of the comments that were received during the Public Comment Period, explain how those comments are reflected in the final Public Patent License (“License”), and describe how other policy considerations shaped the drafting of the License. You may refer to these Notes to help understand our views of the issues presented by the comments and how I attempted to resolve those issues. These Notes are not intended to be part of the License and may not be used to change or add additional terms to the License, although they may be helpful in interpretation of the License when there is ambiguity. These Notes are not intended to provide legal advice or substitute for legal counsel. However, you may wish to consider them, with the aid of your legal counsel, prior to using the License.

The commentary is arranged in an order corresponding to the sections that appear in the License. In order to protect the anonymity of commentators, I do not identify the source of particular comments, and in some cases, multiple comments are summarized together for ease of reference. Some comments that resulted in minor or clarifying changes to the draft of the License during the Public Comment Period are not discussed in detail here, but are reflected in the Change Log. You do not have to review the Change Log in order to review the final License, but if you have been following the drafting process or if you have already reviewed an earlier version of the License, you may wish to refer to the Change Log for a quick summary of changes made leading to the final draft.

Title

A number of commentators suggested changing the title of the License. One commentator expressed the concern that the name “Model Patent License” implies that the License has broader applicability or acceptance than is warranted or that it is more specifically suitable than other patent licenses that already exist. This was not our intention. The current working title is “CC Public Patent License.”

Preamble

Some commentators expressed concern with the statement in the preamble that this License is intended for users who are maintaining their patents primarily for defensive purposes. They point out that in some cases, a public offer to license a patent may reduce the effectiveness or deterrent effect of such defensive use by establishing a presumptive measure (cap) on economic damages that may be proven in the event of infringement. They suggest that high-value or core business patents, even if held defensively, should be reserved. I think it is best that you consult with your attorney on the most effective patent licensing and portfolio management strategy for your unique situation, and I do not intend to recommend a particular patent licensing approach for every situation. Accordingly, I have deleted the reference to defensive use of patents.

Some commentators also suggested that our comment regarding the “offer conditions” are too narrow. For example, they suggested that an offeror could specify that those parties who are currently in patent litigation with the offeror or who have infringed the offeror’s patents be excluded from the offer. Otherwise, it is pointed out, an infringer could wait until its infringement is discovered and then take a license, with the result that it would only be liable for past infringement but it could avoid an injunction against future infringement by accepting a license. I agree with this and so added an additional sentence that the offer may be so limited.

Definitions

One of the issues that arose early in the drafting process was what type of model to use for identifying those patents that are being licensed. The simplest model is enumeration—a list of patents that are being offered for license. However, one concern expressed by several commentators was that this would be too narrow: there might be daughter applications such as continuations, continuations-in-part, and divisionals that issue later and that may not be readily discoverable by a licensee through due diligence. For this reason, we added continuations, continuations-in-part, divisionals, foreign counterparts, etc. to the definition of “Licensed Patent Rights” to supplement the enumeration approach. Nevertheless, a related issue arose with respect to so-called “grandparent” applications or prior dominant patents that may block the use of the licensed patents. For example, a licensor may license an improvement patent without including the dominant patent required to practice that improvement patent. In that case, an unwary licensee may wrongly conclude that it is obtaining all rights needed from the licensor to practice the invention, only to be confronted with an infringement claim from the licensor later for a dominant patent. Some courts have addressed the inequitable effects of this type of problem by finding implied licenses. However, for drafting purposes, I did not want to rely on implied license theories.

One suggestion was to include grandparent applications, just like daughter applications. However, this has some interpretive problems, insofar as the claims of a grandparent application cannot be directly related to the claims of daughter applications using clear rules such as priority, because the priority rules allow only the claims of a daughter application to claim back through the grandparent application’s specifications, not its own claims. This imperfect mapping would produce ambiguities in terms of what is being licensed, and I want to avoid any ambiguity that would result in future undesirable surprises for licensees and licensors. The virtues of clarity and predictability argue against such an approach.

Other commentators suggested that we should not overstate the problem about grandparent applications or dominant patents because unlike daughter applications, which may be yet unpublished, grandparent applications and dominant patents are discoverable by the licensee with a modest degree of due diligence, since they will either have been published or would be referenced by the licensed patents. While this is true of transactions between sophisticated licensors and licensees, there may be other situations where the licensee lacks the sophistication or resources to conduct adequate patent due diligence.

Therefore, a second alternative licensing approach is needed for those situations where the licensor wants to assure the licensee that it has licensed all claims necessary to practice (and that are owned or controlled by the licensor). Under this approach, the licensor designates a “Specification” that it wishes to license—this could be a technical disclosure, a patent specification, or a technical standard, for example—and that all claims covering that specification will be licensed as “Necessary Patent Claims”. To prevent “scope creep” caused by an overly expansive interpretation of the Specification, a clarification is made that only claims that are taught by the Specification are licensed and not merely enabling claims that are not the principle subject matter of the specification. (For example, a purely software specification does not include computer hardware claims merely because a computer is eventually needed to run the software. However, if a Specification has both specific hardware and software teachings, then both types of claims would be within scope).

With these two alternative ways of licensing patents, a licensor may either choose to be very specific with the patents it wishes to license by taking the enumeration approach (in which case, the licensee should be warned to conduct thorough patent diligence prior to accepting the license or practicing the invention), or it may seek to assure the licensee that it has indeed licensed all necessary claims by taking the “Specification” route. Thus, these two approaches should be seen as representing two ways of allocating the due diligence burden between the licensor and licensee.

Several commentators suggested that we added more specific branding requirements (similar to attribution requirements) for licensed products or services. Other commentators argued that it is unusual for licensees to agree to branding requirements as part of a patent license. Yet others argued that licensors would not necessarily want to impose branding requirements (which would involve granting a trademark license) unless it has some assurance of quality control. A trademark license without adequate policing (sometimes called a “naked license”) can result in loss of a trademark right. Furthermore, these commentators point out that trademark licensing accompanied by certain forms of technical guidance or assistance can result in a finding of a “franchise” in some states. Franchises are generally trademark licenses that are accompanied by consideration and a business plan (here, the patent license itself and associated know-how may supply the “business plan,” which is sometimes interpreted broadly). Franchise rules are complex, may require state-by-state registration formalities, and carry penalties including rescission and treble damages for failure to follow formalities. Due to these considerations, I consider a “branding” clause to be a “high risk” clause, and therefore I will leave open the possibility of implementing it through the “Additional Clauses” (see below), rather than providing a standard default.

Grant of License

2.1 One commentator suggested adding the phrase “under the Licensed Patent Rights” to make clear that this license is restricted in scope to the enumerated patents. This clarification is useful, so I made the change.

Several commentators suggested that the original paragraph was awkwardly worded and suggested a re-arrangement of clauses to make it flow better. I agreed and made the change. In particular, the phrase “under the Licensed Patent Rights, in the Territory, and during the Term” should be interpreted to apply to the whole paragraph.

2.2 This parallel license grant was added in order to implement the alternative, “Specification” mode of licensing discussed above. While it is technically possible to include both types of license grants (the enumeration approach of 2.1 and the specification approach of 2.2), and they would operate a two separate types of licenses that do not limit each other, for simplicity and ease of reference, it is recommended that the licensor pick one or the other. This can be accomplished simply by choosing whether to fill out the Licensed Patent Rights list or the Specification in the License Data Record (but not both). If the licensor fills both fields, then it will be granting two separate licenses, but the enumerated list does not operate as an additional limitation on the claims licensed under the Specification.

2.3 One commentator suggested that the ability to limit the “have made” right be extended to each of the other rights under patent. However, due to the unusual and complex nature of such limitations, I decided to leave this sentence unchanged and recommend that such additional limitations, if required, be implemented via the “Additional Clauses” mechanism.

2.4 One commentator suggested that the treatment of exclusive patent rights is centered on U.S. law. For example, under U.K. law, the terms “disposal” and “offer to dispose of” are used in place of “sale” and “offer for sale.” However, rather than extend the list of exclusive rights terminology to include variations from every country, I have opted to make the clarification that these rights include their foreign equivalents in order to generalize to all jurisdictions.

4. Reports

A number of commentators questioned why any reports would be necessary for non-royalty licenses. Other commentators expressed concerns that only low-value or insignificant patents would be offered under this license, limiting its impact. In order to provide empirical evidence for impact, licensors are encouraged to work with licensees to collect statistics and empirical measures of impact so that the impact of offering public patent licenses through clearinghouses can be measured and reported. In order to accomplish this goal, there must be an obligation on licensees to report impact statistics back to the licensor, and this section provides a mechanism to do that. Thus, even in non-royalty bearing licenses, some type of report is contemplated in order to provide the necessary usage statistics, although the contents of such a report shall be at the discretion of the licensor.

Some commentators expressed concern that the data reported by the licensee may contain confidential business information or plans. This section contemplates that only publicly available information will be published by the licensor as impact statistics, or if confidential information is involved, that the licensor would work with the licensee to remove any confidential information prior to publishing any statistics, thus reporting only anonymous or aggregated data that does not identify the licensee if possible. The level of aggregation or de-identification will depend on the specific circumstances and the sensitivity of the data provided. In addition, licensees and licensors may wish to enter into supplementary confidentiality agreements when warranted regarding treatment of sensitive business information. This section should be interpreted as requiring the licensee to provide at least publicly available data and providing the licensor with reasonable cooperation in identifying data that could be useful for documenting public or social impact.

Some commentators asked whether the licensee or the licensor will create use case studies to highlight the impact of patent licensing. Both licensees and licensors may publish such studies. However, at a minimum, it is contemplated that the licensee will cooperate with the licensor so that the licensor may publish such impact studies, if the licensor so chooses. Nothing prevents both parties from independently publishing their own use cases and studies. However, the parties should keep in mind their obligations not to imply endorsement or make other use of the other party’s trademarks under Section 8.

5. Term

A number of commentators suggested that the Term of the agreement be made flexible in order to accommodate expiration of the license before the expiration of the patents themselves. Other commentators, however, expressed the concern that this could cause a trap for the unwary, causing some licensees to rely on a license that has a date certain expiration without taking this into account. I think that this concern would be lessened if the fact that the license expires before the patents is specifically called out. As a result, I added the qualification “Unless otherwise specified,” in order to leave open the possibility of future support for date certain expirations. For now, neither the License Data Record nor the metadata will specifically support such early expiration. However, it is possible to implement such a condition via the “Additional Clauses” mechanism discussed below.

Some commentators pointed out that 11 U.S.C. § 365(e)(1) prevents a licensor from automatically terminating a license in the event of the licensee’s bankruptcy. This would invalidate the third paragraph of this section. Note that this paragraph has been qualified by “to the extent permitted by applicable law.” This qualification saves this clause from invalidity (where contrary to applicable law) but cannot make it enforceable. Therefore, a licensor is cautioned against relying on the enforceability of this provision in all jurisdictions. When this is an important issue, further research on country-specific bankruptcy laws is necessary.

Commentators disagreed over the inclusion of clause (a) of paragraph 4 (defensive termination). Clause (a) reads “unless (a) such claim is first asserted after Licensor has asserted a claim of patent infringement against Licensee or any third party with respect to products or services of Licensee.” Those who support the inclusion of this clause argue that defensive use of patents by the licensee should be allowed, without triggering termination rights of the licensor, when it is asserted in response to a licensor’s first offensive use of a patent unrelated to the license. This fits best with their intuition of what “defensive” termination means. Others, however, opined that it is customary that once patent litigation has been initiated, whether by the licensee or licensor, the licensor should have the option of terminating all licenses, whether or not such license is related to the patents that are the subject of the suit, and regardless of who initiated the lawsuit, so that the licensor may make maximum use of its patent portfolio for litigation purposes.

Given that my goal is to strike a fair balance between the interests of the licensor and licensee, I have to consider the effect of removing this clause. If removed, it would allow the licensor to terminate the license whenever the licensee asserts a patent-based counter-claim, even if the licensor first sued the licensee on an unrelated patent. While it is not my intention to preclude the licensor from pursuing other patent claims against the licensee, if it chooses to do so, it should not be able to also “pile on” patent rights that it has already licensed if the licensee does what it normally must in order to seek a compromise—namely, to assert those patents that it has maintained defensively by way of counterclaim. To do otherwise would be to make licensing a high-risk activity for licensee, that is, either give up its right to assert its patents defensively in the event that the licensor chooses to pursue a patent lawsuit against it, or risk losing a patent license that it has relied upon to build and market products and possibly face an injunction. Such a dilemma would not be consistent with our goal of providing a dependable license that can be relied upon to make investment and technology adoption decisions.

A few commentators suggested that if the licensee terminates the license for convenience, it should not have the right to take a new license under a general public patent license offer unless the licensor specifically agrees. This is in order to avoid certain types of strategic behavior intended to avoid paying maintenance fees or other types of annual fees. I agree with this and have made the corresponding change to the license.

6. Warranties

Some commentators noted the absence of any licensee warranties. An early version of the draft did contain licensee warranties, but other commentators suggested that those warranties (such as compliance with law and applicable regulations) were best dealt with through indemnification, because violation of such warranties should not necessarily lead to license termination provided that the licensee fully carries out its obligations to indemnify. I agreed with this and thus removed licensee warranties.

Other commentators suggested that the licensor should provide warranties that it has licensed all necessary rights or that it is unaware of any infringement. However, opposing commentators noted that such warranties are very unusual, and it is customary for a patent license to assume that the license itself does not give any assurance of the right to practice (i.e., freedom from infringement of dominant patents). Furthermore, issues regarding the possible existence of the licensor’s own dominant or related patents have been discussed above and need not be dealt with through the warranty section (or through breach of warranty remedies, which would create a chilling effect on licensor participation).

7. Indemnification

Some commentators suggested making the indemnification provision mutual (i.e., provide for an indemnification right for the licensee as well as licensor). However, this is very unusual and goes against industry custom, in which the licensor’s primary “obligation” under a patent license is to refrain from suing the licensee for infringement. Typically, the licensor does not warrant that its rights are valid or that it will take other affirmative duties. Therefore, there is very little basis for the licensor indemnifying a licensee.

Several commentators noted with approval the right of the licensor to take over the defense of any claim (at its own expense) that involves an adjudication of the licensor’s patent rights. Given the importance of these rights, most licensors would not want to leave their defense or settlement in the hands of a non-exclusive licensee.

8. Trademark and Attribution

One commentator noted that some patent owners would not want their non-exclusive licensees to be marking products with patent markings without their permission. Therefore, this should be made at the option of the licensor, but not automatic. I agreed with this and added a corresponding qualification: “At Licensor’s request”.

One issue that commentators requested clarification about was whether the licensor or the licensee would be preparing the use cases that document the impact of the licensed patents. As discussed above, while there is nothing to prevent both parties from creating and publishing their own use cases, the intention here is require the licensee to provide reasonable data and statistics that the licensor would need in order to compile use cases and statistics that provide empirical evidence for the impact of its licensing strategy. Therefore, this provision contemplates that the licensor will be preparing a use case with data and support from the licensee. Once published, such use cases may be distributed by both parties. However, aside from the distribution of such case studies in unmodified form, neither party may use the trademarks of the other or otherwise imply an endorsement by the other. No general trademark license is implied.

9. General

A number of commentators suggested that the insurance requirement be made more specific. However, as the size of licensees and the extent, nature, risks, and markets of the possible exploitation of patent rights vary considerably, there is no “one-size fits all” approach to determining the appropriate amount of insurance. Thus, I define it at a “prudent” level so that the above factors can be taken into account.

10. Additional Clauses

I added this clause in order to anticipate future extensions of the license through CC-supported modular clauses as well as to make it possible to use the core license with unsupported “custom” clauses, when the licensor chooses to do so.

The CC-supported use case involves adding additional clauses in the future that we may discuss with the community and that we are convinced will make the license better or add additional options for which there is community demand. Such clauses could include, for example, date-certain expiration of the Term, options for additional limitations to the license scope, or alternative reporting requirements. Such supported additional clauses would be added after consultation with our board and our community through a public feedback process.

The unsupported use case involves a licensor who wishes to implement a feature that we do not currently support in our standard license. These could include certain “high-risk” clauses such as branding and grant-backs, that a licensor may wish to implement only after careful consultation with its attorneys or other types of clauses that CC chooses not to support.

For example, I have already discussed above why branding and trademark license clauses should be reviewed carefully in light of trademark licensing practices, policies, and franchise law. Due to the complex nature of this analysis, it may be appropriate for some licensors to provide such branding or trademark licenses, while doing so may raise serious problems for other licensors. Creative Commons would not want to implement this as check-the-box option, because it poses a risk about which some licensors may not be aware. On the other hand, a licensor who chooses to draft their own clause and implement it through the Additional Clauses mechanism is more likely to have considered the issue carefully and is prepared to take the risks involved.

Another example of such a clause is a grant back. Several commentators have suggested that we implement some kind of grant back. Some have suggested that we implement a type of “share alike” for patents that would require licensees to grant rights in improvement patents under the same license, whether to the original licensor or to the community as a whole. Prior experiments with such share-alike licensing mechanisms in a community licensing context (or patent pools) have met with decidedly mixed results. Nevertheless, a grant-back is a well-known and accepted practice in traditional patent licenses in order to ensure that the licensor has access to improvements and freedom to operate. However, the use of grant back clauses carries some antitrust risk. The 1995 Federal Antitrust Guidelines for Licensing Intellectual Property recognizes that certain grant-backs can pose antitrust concerns where the licensor has market power. Grant backs are analyzed under a rule of reason standard, and courts have also found some grant backs to lead to antitrust violations (i.e., the use of grant back clauses by licensors having market power to achieve a monopoly) (see, e.g., U.S. vs. Alumina and U.S. vs. General Electric). Compulsory grant back clauses raise antitrust concerns because they may reduce the licensee’s incentive innovate relating to improvements. Thus, grant backs are appropriate for some licensors but no necessarily “safe” for everybody, and they deserve special analysis and review by appropriate antitrust counsel. For this reason, I decline to implement a standard check-the-box grant back option, but I leave open the possibility that a licensor may supply such an obligation through the Additional Clauses mechanism.

Licensees should also be alert whenever there are Additional Clauses that are not supported or endorsed by Creative Commons that such clause may require special scrutiny because they have not been reviewed by Creative Common or subjected to public comment, and licenses which contain unsupported Additional Clauses may not be branded or labeled as standard CC licenses.