Jason Schultz spoke about creating a kind of GPL for patents (and then my computer crashed, so sad). The project is designed to deal with one of the big barriers to defensive patenting: the fear in open source types that a defensive strategy will not stay defensive once a property right exists, given that plenty of companies have said that they were only patenting defensively and have later changed their practices. So the idea is a nearly irrevocable license to anyone who agrees to reciprocate across their entire patent portfolio; if you then start using patents offensively against people in the GPL community, you lose the license.
Dan Burk: Questions about antitrust.
Van Houweling: Some hesitance about licensing generally—Niva Elkin-Koren has noted that pre-CC was a world of lots of tolerated uses; CC licenses may encourage people to think more about themselves as rightsholders, and there are possible inconsistencies between licenses. Claiming of rights by user-innovators can raise transactions costs, creating a bigger mess and bigger need for exemptions. Which isn’t to say that exclusion rights are unjustified, but using exclusion rights to create openness is tricky because of difficulties communicating about the terms and can create further barriers to innovation in ways that relying on the public domain wouldn’t—especially given that there is an automatic public domain in patent as there is not in copyright. What about publishing instead? Or what about trusteeship—transferring to some trusted intermediary?
Carroll: standardization around a license has real implications—OpenCourseware is finding that interoperability is emerging as a more important issue as a community begins to see itself as pulling from and contributing back to a larger pool. So the open courseware community has standardized around sharealike noncommercial, but attribution-only (or attribution-noncommercial) might be preferable; it’s a big debate and incompatibility issues exist. Nike approached Science Commons for Green licensing where Nike would license potentially useful environmental technology to noncompetitors/other industrial applications; Nike has put 100+ patents into this.
Stephen Maurer: There are people who have an urge to donate and you need to keep them involved. Synthetic biology: network effects around knowledge—better to have it held by a community rather than a private company. First best for company is to have access without donation; there you need to use IP or some other vehicle to enforce the second-best, in which the company contributes to and draws from the pool. That may allow the community pool to grow faster than the privatized pool, producing network effects without monopoly. Donation model and ruthless commercial trade are both open source, but very different from a design perspective.
Strandburg: has anyone looked at the endowment effect with respect to ownership (Chris Sprigman and Christopher Buccafusco have a haiku experiment) of patents/IP in this sphere? How do you bring people who don’t want patents into the sphere? Could we also say that the license covers people who don’t have patents?
Schultz: You can take the license no matter what you own or don’t own.
Strandburg: what if everyone disavows patents and then you have no defense against the patent troll? How much worry is there about the big outside bully?
Scotchmer: Two views of open source licensing. One, romantic view of contributors to common good; other, cold-blooded business strategy—these are reachthrough licenses, which are ordinarily considered nefarious and not common-good-promoting. If the first view were the only view, then dedication to the public domain would be sufficient. But open source can be collusive: agreement not to innovate and assert control in a rivalrous way. Open source can also be profitable because we agree not to hold each other up—there, there is no conflict with the public interest. Is this just a way for big firms to profit from little firms? If the locus of profit is this unprotectable learning-by-doing; if you’ll profit from your knowledge regardless of whether it’s patentable by getting little guys to contribute to making applications for your platform, then it’s just good business.
Von Hippel: for user innovators who benefit from the act of inventing—it makes them feel good about succeeding at something—that’s not dependent on the behavior of others; you want exemptions, not (necessarily) rights.
Baldwin: this form of property rights, in the hands of some players, can be an instrument of collusion to not invest in innovation. Those players wouldn’t be classic users, because for users the value of innovation is use value, so they always have motivation to invest to create value in use. What the collaborative institution does is allow splitting of effort and recombination of artifacts into larger systems that benefit nonrivalrous users. It’s not a desire to donate so much as a use-profit-seeking effect.
Maurer: open empirical question in synthetic biology: what’s the composition of the group? People at universities get paid anyway; but there is also a lot of money available commercially. If your big concern is that commercial entities will surprise us with a monopoly on some important technology, then the choice of policies is different.
Von Hippel: why isn’t an exemption just as good for users as a right?
Samuelson: there’s more than one way to achieve exemption-like status. Some policy levers like exemption might not be on the table right now, so don’t throw away rights as a possibility of achieving the desired result.
Baldwin: question: who’s the user core? Are there people who just like designing/get joy out of belonging to the community? Reduces costs—these are people for whom design cost is low. Other people: commercial participants, who find it better to be part of the community than outside it. Empirical question about who dominates. Suspicion: viable versions of this model have users who really care about the thing itself at the core, and have a hunger for design improvement. Nonuser/commercial groups will be attracted to a vibrant, healthy institution, but can’t take it over or it will die.
Kapczynski: because copyright is so hard to get rid of and patent so hard to get, open licensing will be more useful to the former. What is the proper form of the license? Folks who are not interested in commercializing may not be interested in resisting others’ commercialization with reachthrough licenses. What counts as open for them? Other conversation: for people who do open licensing, what is necessary to make space for users? Royalty-free versus RAND (reasonable and nondiscriminatory terms).
Dietmar Harhoff: Skeptical that general user exemption without restrictions on what “use” means is feasible. German patent law insulates prior users who don’t file for patents from later holdup by people who do get patents—let people stay out of the system and be safe.
Van Houweling: patent licensing may be less satisfactory because of the difference between copying and independent creation with respect to liability. When you have a copyright license, you know what you can do, but when you have a patent license, you don’t know very much; might be other patents out there. (On the other hand, even with a copyright license, you are still vulnerable to claims by some third party that you infringed their copyright—consider Coming to America, where the studio clearly had licenses from most of the relevant people.)
Session IV (Intermediaries and Secondary Liability)
Discussion of Blizzard case: the infringement theory was that the RAM copy of a user’s game was unauthorized because, though the user was an authorized user, the particular use was unauthorized.
Dan Burk: isn’t this what the GPL wants?
Schultz: Federal Circuit has now interpreted Grokster more than any other circuit—inducement is an issue in many IP areas, though Grokster may not have influenced the treatment of patent much. Inducement continues to be a big deal—recent Limewire case—so it chills people who are trying to help innovators. Level of knowledge accepted to trigger in copyright is much more general and vague than in patent—general knowledge that users are infringing something seems to suffice, whereas in patent you’d have to know about the specific patent that the users are infringing.
Notes that Facebook has sued Power for extracting information to aggregate all their social networking data; Facebook claims to own all user information, including photographs, status updates, and so on.
Session V (Infringement Exemptions, Fair Use, and Exhaustion)
Patent Act §287(c)(1): methods of surgery are patentable, but not enforceable against doctors, helping personnel, or institutions in which they’re done—Dan Burk says it’s a complicated and unclear provision. Not sure we understand who’s covered.
Strandburg: Only one case interpreting it, involving an independent lab, finding them covered by the exception.
Maurer: Secondary liability is the purpose of retaining patentability—someone who gives you an instrument to perform the method may be liable.
Samuelson: arose out of a case in which one surgeon sued another; the community went nuts. Debate over patentability v. exemption for practice. Partly because defining exclusion from subject matter was difficult, and because of concern about covering methods that doctors don’t actually practice, the decision was to create exclusion around people practicing medicine; also a sense that this would be more compatible with treaty obligations. Drugs aren’t covered.
Strandburg: user innovator community had enough clout to get Congress to pass a law for their social norms. That may be why there aren’t many cases—very few people would flout the norm. Doctors have the same reaction to diagnostic method patents, but people outside the community are involved in those; will they be able to get the same protections against those patents? There, absent further congressional action, relief would have to come from patentability rather than a practice community exemption.
Schultz: Boundaries of community: users are usually thought of as amateurs, but these are professionals with certification/licensing and a national association representing them. These boundaries and accompanying social status may make the exception seem more tolerable—it can’t expand.
Kapczynski: And the bounds of that community map pretty well onto the bounds of the community that would be engaged in the innovation in the first place.
Strandburg: which makes diagnostic tests an interesting comparator, where innovation may also come from outside though it’s not clear how often.
Dietmar Harhoff: The harm of these patents is in the overall information structure—becomes harder to determine whether you are violating any rights. Patent examiners think that these patents will be narrow, but their harm comes further down the line.
Samuelson: in theory, you could still get a patent, but with luck people will be deterred from patenting because they can’t enforce it against the key players.
Harhoff: but there are lots of reasons for patenting; not entirely protective.
Carroll: Tailoring can work on scope, on defenses, or on subject matter. Tailoring on subject matter requires confidence that what you’re carving out really doesn’t belong in the system; EU has done that but Supreme Court is unlikely to.
Harhoff: Proposition: Small patents clog up the patent system by creating uncertainty.
Samuelson: but doctors are unlikely to worry about patents with the exemption in play.
Burk: The argument is that trolls can succeed by going after small/medium size players and asking for sums less than the cost of litigating, even with invalid patents. Scope reforms won’t help with this.
Strandburg: likes use exemptions because it’s easier to figure out what a user is doing than to figure out whether a patent is “really” a business method.
Kapczynski: Ways to make a user exemption—a status? Not plausible, because no status exists that covers “users” in general. Motivation? Noncommercial—then we have to define that; may differ between patent and copyright. Form of the use? Also a relevance to space—private use?
Burk: only have one real subject matter carveout in the US, nuclear weapons. Compulsory license for civilian nuclear tech—technology-based rather than user-based; Clean Air Act has similar provisions. Could consider remedy limits to be a type of exemption/carveout—injunctive relief might be denied for certain uses, e.g., sewage treatment. Exemption for generic drug development.
Schultz: look at §117 in copyright—a right to tinker if you own a copy (from a period where that kind of tinkering was a more standard part of working with a computer program); ownership of a copy of software is uncommon now given licensing developments, so that’s not suitable but it’s a place to start.
Strandburg: line between repair and reconstruction is also unclear, especially since very few cases involve users. Possibilities: some kind of justifiable failure to license, like Wendy Gordon’s transaction costs rationale for fair use—market too small, or patentees trying to squelch further innovation, or anticommons type holdups like eBay v. Mercexchange. Added on: “blocking patent failure.” Ordinarily we say that a substantial improvement can be patented, but someone in an open source environment might not want to do that. Other considerations: Industry specificity: where are patents important? Exemptions should be granted more where there is less reliance on patents for innovation. Knowledge: is the user an independent inventor? Would the patent have been easy to foresee/find?
Peter Lee: Legislative history of medical technique exemption: doctors are well-organized, but it was also influential to say that the exemption helped patients. Build a narrative of the people who are harmed by the rights.
Van Houweling: Exhaustion seems both useful and problematic (because of boundary uncertainty and undertheorization). Fiddling with your own stuff as a motive hasn’t been discussed in the courts. Oral argument in Quanta: some Justices are concerned about users and even user-innovators.
Schultz: making and using v. importing and selling?
Burk: almost no cases about what “use” means in the Patent Act. Early cases suggest that “use” is pretty broad, but don’t give many reasons. Stockpiling ammunition: court says storage in preparation for war is use.
Carroll: courts have read a volition requirement into copyright—nothing in the statute says that but it’s a policy-based reading. Stories about consequences could lead a court to adopt such a reading. Farmer whose field is accidentally contaminated with genetically altered seeds might be an example in patent.
Strandburg: very creative interpretations of “use” exist in prior art context, which covers “public use.” “Experimental use” too—shouldn’t we try to interpret these terms in some coherent way.
Brian Carver: We’ve been patent focused for a while. Lots of user creativity going on in copyright too. Should that be infringing? There are albums that are classic and game-changing that couldn’t be made in the legal climate we have now: tell those stories.
Van Houweling: Rob Merges would say that these examples of user innovation show that the system is working, no chilling effect. Stories about things that can’t happen now can help.
Von Hippel: so the law is incompetent?
Van Houweling: lots of things fall through the cracks; she finds compelling the argument that we shouldn’t have to operate under the sword of Damocles.
Samuelson: Tim Wu’s argument about tolerated uses—things that are tolerated become accepted, putting burden on rightsholder to raise an objection that may not be validated by courts. Zone of tolerated use becomes normative.
Me: I don’t like the concept of toleration. Toleration is about power: toleration is about the normative allowing the non-normative to exist. Toleration can be withdrawn.
Van Houweling: though in Wu’s version it can’t be withdrawn if it lasts long enough.
Burk: laches, estoppel, adverse possession—toleration can have permanent effects in law. (But that has to be individual and the idea Wu presents is more that if owners generally tolerate something than a particular owner shouldn’t be allowed to object to the practice applied to the owner’s works; this is a massive and I think unlikely revision of the usual treatment of toleration in the law. Compare how hard it is for the public to get an easement over your land v. a neighbor.)
Scotchmer: We have a concept of equal protection; there’s no equal toleration under the law. People who get tolerated tend to be people with power: who gets tolerated when they park illegally? Who gets tolerated when they engage in illegal construction on the block? Some people get deference and others don’t.
Strandburg: can think of tolerated use as static or as an input to strategy. This may all be time-dependent. In the beginning, if someone’s the first person to fight back, they might lose. At a certain point, we find a person who can win. That comes up in gay rights too (comparing concepts of “toleration”)—there comes a point when it would look really bad to shut people down.
Schultz: Operationalizing licensing: some machinima licenses allow noncommercial uses—space of safety for some creativity, but it’s still a benevolent dictator. AU’s Best Practices around fair use—also trying to find particular genres of creativity, not with licensing but with naming norms around which to galvanize.
Samuelson: private use exemptions as exist in some countries might be another tactic in this space; US has tended to bundle with fair use, but we don’t need to. Could also tie limitations to definitions of exclusive rights—public performance right = implicit license to everyone to make private use. Or require commercial harm as a prerequisite for finding a violation of the excusive right.
Schultz: IP counsel worry about “sitting on their rights”—this generates a lot of bad threats, and we might want to avoid giving them a rhetorical excuse for going after people who aren’t doing any harm. (This is another reason I don’t like “tolerated use.” I think it encourages this type of argument: if I don’t go after it, then I’m tolerating it and may lose my right in its entirety—that’s not true as a matter of descriptive law, but it’s widely misunderstood.)
Gordon: Tolerance and custom: they are related, but custom doesn’t have some of the uglier connotations of custom. Systematizing relations between custom, toleration, and laches—Lloyd Weinreb argued for using custom to increase the scope of fair use. Jim Gibson argues that custom is a bad source of reference because customs of licensing arise out of fear rather than out of right (in either positive or normative sense).
Burk: The law doesn’t require you to go after everyone (I agree, but there’s a lot of misconception surrounding this), but estoppel and acquiescence are doctrines that are there for a reason. You don’t want IP owners to sit on their rights while you make a big investment. There’s still an important function for estoppel; don’t throw it out.
Carroll: Toleration intersects with our concept of the scope of rights: in one concept, the right is far-reaching and the owner may tolerate a bunch of stuff within the scope of the right; in another, the right is limited, and overclaiming is a misuse of the right.
Samuelson (and others): discussion of misuse as an alternative—misuse can work like an exemption during the period before the misuse is cured.
Harhoff: influence of behavioral economics on whether people withhold innovation—there’s now general agreement that patents not be licensed even when licensing would promote progress/efficiency, because private and public interests need not be aligned, but we need now to add in things we know about imperfect rationality (including managerial conflicts).
Samuelson: we should also say more about fair use. Fred von Lohmann argued that should also rely more on first sale to support tinkering. Fair use does a lot of work (being asked to do too much) to facilitate innovation on the copyright side—potential application to patent reform? Can’t expect more general exemptions from any place but the courts these days. Courts have accepted licensing to defeat first sale much more readily in copyright than in patent, but we can push to change that.
Burk: deal with whether we should give people exemptions or rights; rights may allow them something to bargain with.
Session VI (Remedies)
Andrew Torrance: eBay v. Mercexchange has been characterized as a compulsory license regime. Have to treat patents like the rest of law in terms of injunction v. damages; 72% of the time injunction is awarded now as opposed to nearly 100% in the past. Reasonable royalty v. lost profits? Would make it easier on user-innovators in terms of risks faced. Denominator problem in patent reform: if I make a car with one screw covered by patent, what is the basis for calculating damages? Willfulness is also an issue. Also, even if you win, there’s a litigation penalty--$3-5 million to take a patent dispute to judgment.
Burk: Think broadly—restitution instead of damages/injunction. User innovation: the key is the preliminary injunction stage, so we should look at that in particular. Posting a bond: FRCP has provisions for putting the plaintiff at risk too.
Strandburg: allow users a setoff or some other consideration if user innovation had contributed to the invention of the patentee—what would happen naturally if both sides had relevant patents. (Does she mean that the particular user would need to have contributed, because that’s an unlikely scenario? How would we structure asserting third party contributions to the invention?)
Schultz: Need $100,000 to get to the PI stage—a bond might not be much help. Are there ways we can deal with the punishing cost of litigation before that? Remedy for knowing misrepresentation in DMCA notification—might be a way to push back. Question about burden of proof for defenses/exemptions at the PI stage. Defendant pleads fair use as a defense, but fair use is not an infringement of copyright according to the statute. Ninth Circuit in Google v. Perfect 10: court initially put the burden in some ways on Perfect 10 to prove it wasn’t fair use, and then took it out of the amended opinion. Tweaks would be useful. Winning on motion to dismiss: has happened in a couple of cases, and that’s even better.
Torrance: for users, have citations/really low fines for infringement, $100. Users could keep going: it’s a parking ticket.
Kapczynski: International law: are these deterrents?
Harhoff: US litigation cost allocation rule scares the crap out of the little guy; even if he has a good case, the other side can unilaterally drive up costs, and the cost level is extremely high comparatively. Average German patent case: 150,000 euros, over in 2 years, and you get reimbursed if you win; similar in France; 1 million pounds in Britain and resultant higher settlement rate. Which also means that validity of patents is swept under the rug. People have strategic incentives to litigate. Possible damages: the cost of inventing around—this would be better as a deterrent to trolls (Scotchmer points out that this could be reasonable royalty in many circumstances, though the law is messy). There are virtually no trolls in Europe because the cost levels/strategic use of litigation is not as possible. But Europe may be moving in the American direction.
Burk: reasonable royalty is like the Holy Roman Emperor: none of the above. It’s not designed to replace negotation, because the negotiation is only hypothetical and because courts are concerned not to give people an incentive to infringe, knowing that at worst they’ll only have to pay what they would have agreed to pay voluntarily.
Brian Carver: recall that statutory damages are good for users in one place: open-source licensing, when it’s violated. Jacobsen v. Katzer. Open-source projects are generally small and judgment proof, and the thought from some is that they don’t need to worry about patents—but that’s not true because there are commercial entities that want to control the market; if the commercial entity in this case could squash the open source competition, it would be the only one with model train software available. So enforcing the open source license against the commercial entity is a big deal. Turned out that the patent holder had copied open source software and removed the attribution; put the software project in a strong position for settlement. But Jacobsen had to live with a period in which his house was under threat: he’d gotten letters saying that the company’s patent read on his software project and thus he owed $200,000; the company also contacted his employers and did what he thought threatened his livelihood/grants.
Discussion about the class of unlawyered/unresourced innovators: when the complaint shows up is the key time for them. Tweaks later may not help.
Carver: make post-grant review a precondition of filing a patent lawsuit? Might help the little guy; could make this requirement only triggered for noncommercial users or other groups about which we’re worried. This would allow anyone who wanted to make an argument to bring it to the Patent Office’s attention.
Schultz: certain politicians have embraced new kinds of stories about participation/remix culture. Mike Doyle & DJ Girl Talk as a talking point. Rick Boucher: got in the fight because he was interested in libraries and fair use. Innovators in particular districts should talk to the representatives from their districts.