Copyright Office, 1201 Roundtable, DC
James Madison Building, Mumford Room
NB: I won’t be here tomorrow because the Office had the temerity to schedule the roundtables after I committed to a different exiting roundtable, this one at Notre Dame on deception. I wish I could be at both.
Session 1 Relationship of Section 1201 to Copyright Infringement, Consumer Issues, and Competition
This session will explore the role and effectiveness of section 1201 in protecting copyrighted content, and will consider how the statute should accommodate interests that are outside of core copyright concerns.
CO: Jason Sloan, Kevin Amer, Regan Smith, Abi Mosheim
Smith: 1201 is part of DMCA; Congress recognized that TPMs could be deployed not only to prevent piracy but also support new ways of dissemination of © material to users digitally. Protects access controls, trafficking in circumvention tech for access and rights controls. Regulatory process has become burdensome, particularly upon renewal; permanent exceptions may not have adequately foreseen developments.
Amer: big picture: overall role and effectiveness of 1201. Many comments noting the different distribution models TPMs facilitated; other comments questioned the relationship between legal protection of TPMs and effectiveness of TPMs.
Jonathan Band, Library Copyright Alliance: lawyers like to feel we have an impact on the world, but TPMs, to extent they’ve been effective, it’s b/c the TPMs have been technologically effective. Saw early on ineffective measures such as CSS, very easy to hack, and the tech to hack it was widely available. There was some hacking, some not, but we now are generations beyond that, especially w/r/t content in the cloud behind paywalls. The law has nothing to do with the effectiveness of those systems, especially w/r/t the general public. What makes those work is the tech, not the legal protections. Hacking into systems is the purview of the CFAA and state anti-hacking statutes. [Not to mention that many of these threats are foreign and also by dedicated lawbreakers; not deterred by anything we can do.]
Allan Adler Association of American Publishers: Recognized early on [that is, before there was evidence], b/c this legislation was the result of an int’l treaty determining that legal protections were important. Common sense that ability to use locks means nothing w/o legal restrictions against violating locks or creating devices [like secondary liability?]. Compared this to breaking into a locked room to get a copy of a book. [Yeah, that’s totally what vidders do.] Convenient, cut expenses, make works easier to find: but access online had to be secure in the face of all the ways in which online networks are susceptible to unauthorized activities. [There will be no cars on the information superhighway without perfect protection!] Wild success.
Jonathan Zuck, ACT | The App Association: remembers buying Copy2PC, in my younger unruly days—I copied software I wanted to use; it took creating software to respect it, in my case. [Which is why we should make it hard for people to create new stuff via remix …] It’s not the availability that matters, it’s the inability to walk into Best Buy and get the software that matters—practical availability, not binary connection.
Raza Panjwani, Public Knowledge: WIPO treaty thought necessity to provide protections to © owners to encourage them to provide digital goods and distribute them online. We’ve seen DRM unnecessary to various business models—iTunes, game distribution platforms, Tor book distribution. Not necessary; 1201 burdens noninfringing uses to provide protection, and CBA is justified—do we still need high barriers to encourage distribution? We should also ask how effective 1201 has been at discouraging piracy, not just “circumvention”—comments decry how much piracy there is. [Including, not for nothing, Adler for the APA, as I recall from a few weeks ago. Zuck has a point about Best Buy, but the corrollary is that a digital copy in the wild circulates freely once any DRM has been circumvented once. Bittorrent doesn’t require you to walk anywhere.] The content allegedly made available by circumventing DRM and the content existing in marketplaces is the same—don’t conflate correlation with causation.
Q: how do we get at that? 1201 actions: relatively common or not?
Stanley Pierre-Louis, Entertainment Software Association: new consoles, games, are possible b/c of goals of DMCA: the thought was that we want to expand consumer access to broadband and new tech, and encourage that by getting © content online and on devices. We’ve seen that growth, which makes our industry a growing one, $23.5 billion—we know that our investment in our tech and software will get protected. When we’re talking about DRM-free content, that’s a choice; 1201 doesn’t require it. [Unsurprising evasion of the question.]
Q: do you use 1201 in court?
Pierre-Louis: we’ve had a lot of success in preventing access, so we don’t make legal claims; but we know that have that as a backstop. Internationally as well. [Even in places w/o 1201 style law?]
Troy Dow, The Walt Disney Company: There is real value in the tech. The law was also important. DMCA was directly relevant to adoption of CSS, allowed intro of DVD, which was fastest growing device. That was hacked, but that also spawned precedent of DMCA in enforcement litigation. Not to say that these techs are hack proof, but that you keep them out of the mainstream. Zuck talks about Best Buy: when we were in litigation with 321 Studios, that was exactly the issue. 1201 keeps that at the fringes. Business side: availability of legal tools was directly relevant to getting into these markets—ACS, 4k, over the top TV, streaming TV through authentication—1201 a factor in all those things.
George P. Slover, Consumers Union: It’s not the intended uses, it’s the overbroad unintended/uninvisioned uses that are a problem. “Breaking into a house to steal a book”: compare “my wife has baked a cake she intends to take to church; she can tell her hungry son not to touch the cake on the counter or she can lock him out of the entire kitchen.” Reconfigure to focus on the protection against infringement, not interoperability and less closely related side benefits to industry of denying access.
Q: Dow said 1201 successful for motion pictures. Has 1201 played a useful role in the markets for 3D printers or cars?
Aaron Lowe, Auto Care Association: Aftermarket is a huge industry in the US. Use of software in virtually every component. We can’t tell if it’s functional or simply to make it difficult to make aftermarket parts. 1201 is used to keep our industry from reverse engineering and even servicing cars. We see more and more parts with chips on them making it difficult for aftermarket to exist.
Brian Weissenberg, Institute of Scrap Recycling Industries, Inc.: Exactly—last year, when we asked for phone unlocking, phone co. said we rely on DRM to protect our business model, but that’s not a © interest.
Zuck: Non-DRM business models do exist; the people that make use of DRM don’t like it. It increases consumer support costs, but it’s a necessary antipiracy tool. Successful in software market; distinction b/t software and content is increasingly blurred. App that teaches yoga positions: sometimes piracy is pulling content out of app and repurposing it. Final point: expanding beyond intention—I’m not a lawyer, but my understanding is that where adjudication occurred they’ve interpreted the DMCA more narrowly for garage door openers and printers [unless you’re in the 9th Cir.]. The system worked—jailbreaking exception from Congress. Be cautious about upending success.
Band: There have been relatively few 1201 cases, especially relative to 512. Keep in mind its role in larger construct. One reason why so few is that it’s really broadly drafted; talks about active circumvention and 1201(b) trafficking; early cases were strongly in favor of rightsholders; there hasn’t been a need for litigation because it’s so broad. Then the action has shifted to the rulemaking b/c of all these adversely affected people. You have 2 choices: go ahead and hope that in the event of litigation you’ll end up in the Fed. Cir. or the 6th Cir. and not in the 9th Circuit. Good luck on ensuring that! Or you can go to the rulemaking. That’s where the energy and activity has all shifted. That’s why so many of us are so intent on getting it to work better.
Robyn Greene, New America’s Open Technology Institute: Agrees with Band re: not much litigation. It’s impossible to calculate the chilling effect on the market. Entrepreneurs trying to enter a marketplace don’t want to take the chance that all their efforts, expenses etc. are all for naught even if they eventually win litigation. Litigation has been pushed to rulemaking, and that’s still a chill, especially for activities very clearly unrelated to infringement, such as security research, due to the burdens of the exemption process.
Q: add an infringement nexus to 1201?
Adler: there’s already an infringement nexus, but not in 1201(a); access is not part of exclusive rights under 106, but that’s the threshold issue w/r/t uses that implicate exclusive rights will occur. Congress structured 1201(a) to deal w/ access independent of the issue of infringement b/c access was whether consumers could exploit the marketplace for these work and whether there’d be a market to exploit. 1201(b) doesn’t address circumvention of TPMs that protects rights of a copyright owner b/c it recognizes those rights are themselves subjects to exceptions like fair use—have to look at questions differently. Access isn’t about infringement. [Which is why (c) owners have merged access and rights controls.] 1201(b) is where the balance w/exceptions and limitations were struck. [But merger destroyed that balance, and you guys did that, not us.]
Band: yes, but Congress was thinking about access to something you hadn’t paid for, such as access to premium cable you hadn’t paid for. Not the situation that Lowe deals w/in auto context, a person not being able to access her own software in her own car to make sure she doesn’t pay marked up prices for repair parts. No policy reason for restrictions on rights to access their own copies for which they have paid. That’s why we think changes make sense. Whether through statutory change or liberal application of the rulemaking in the case of at least software where we see this problem most pervasively, we could simplify the CO’s work considerably and not regulate the entire US economy.
Q: is your theory dependent on owning the good, rather than renting? Fear that consumers may have paid for a penny but want to access a pound? Does it matter whether you own the refrigerator or just lease it? Is it different w/ a book?
Band: there are various ways to slice the loaf; if amending 1201(a), the best way would be requiring a nexus to infringement. [The only reason fridge makers now assert that consumers aren’t really fridge owners is the legal exploitability created by 1201. It was uncontroversial that your thousands of dollars bought you a fridge until then; they don’t take it back if you breach your “license”; the legal regime is driving the ownership characterization and that’s bad.]
Q: would nexus encompass the paradigm “access to watch”?
Adler: problem w/requiring infringement nexus is fails to recognize that access has its own independent value w/o regard to whether someone will infringe. We hear all the time about print v. digital that reading a book isn’t something that exploits one of the exclusive rights of a © owner. True, but purpose of putting a book on the market is ultimately to have people read it. [Won’t that involve copies, digitally speaking, especially given the RAM copy doctrine that Adler surely endorses?] Only way to control access is to not require infringement nexus.
Band: could be drafted; if I rent a book for a week, and I fiddle w/software so I can keep it another week, that might get into what we’re actually interested in, especially since “reading” involves RAM copies of more than transitory duration. Even in that case, it could be drafted to be w/in the scope of amended 1201.
Dow: Adler is right; absence of nexus to infringement is deliberate; proposals to require nexus were rejected b/c that would make it duplicative of 106. What Congress saw was a world in which value to consumer and © owner was derived from access, not 106 rights. Doesn’t require copying, distribution, public performance, but just access: celestial jukebox. Congress had that in mind. [But not the celestial fridge!]
Panjwani: Nexus could help a lot. “Exclusive right of access for authors” is the claim of Congress’ intent. So: Let’s have a debate about right of access, not circumvention. If Congress had wanted to do that, they could have put it in 106. Instead, Congress said it wanted incentives to enter digital markets, and thought this was a necessary step. Let’s now evaluate that. WIPO treaty says adequate protections for TPMs employed by authors in the protection of their existing rights. Adler/Dow: underestimate ability of plaintiffs’ counsel to find 106 violation: lawfully possessed after the rental expired? Unlikely! Doesn’t require such a broad prohibition, burdening all other noninfringing uses.
Q: speak to distinction between access and copy controls in 1201(a) and (b).
Panjwani: Antitrafficking provisions’ distinction in implementation has not worked. Access control and copy control are the same in practice.
Q: b/c they’re merged in industry?
Panjwani: yes, in practice.
Pierre-Louis: in our industry that’s not the case. Publishers and disk manufacturers have separate types of access and copy controls. We’re not trying to ID infringers, but cultivate consumers. Access to work helps distinguish users—some want it mobile, some online, some on disk. These rules let us meet customers where they are. Nexus to infringement harms the consumer, b/c right now possession of the work isn’t infringing, it’s the uses. That access control allows them to make uses.
Slover: Ownership v. rental: Ownership is the core focus. The rights of a consumer who owns a product to use it. Be careful: ownership can be written around by lawyers so a consumer thinks she owns something but doesn’t. Concept: consumer’s rights and dominion about the product they paid for. It’s a useful starting point to clear the smoke around the tech. Traditional incidents of ownership should matter. Auto software: if consumer could do it in garage w/screwdriver in the past, it wasn’t a © issue. In the new world, may require access to software to make adjustment; the fact there’s a TPM on top shouldn’t change the ultimate calculus.
Q: software embedded devices study: could we treat those TPMs differently than TPMs protecting expressive works, instead of infringement nexus?
Slover: not a © lawyer, and Consumers Union came into this from phone unlocking. Can’t speak too definitively, but sees a big picture conceptual distinction b/t core creative works like books, movies, songs and products that you now can’t use unless you have software inside them.
Adler: Panjwani suggested that Congress could have included access w/in exclusive rights of © owner. Presumably his constituents would vigorously oppose that, b/c it would make access a property right as the other aspects of 106 are [hah!] as opposed to encouraging use in a market and ability to assert terms & conditions of availability & use by controlling access in the first instance. Concerned by emphasis in notice on 1201 “outside of core copyright concerns.” Not sure what that means, but mistake to equate that w/notion of nexus to infringement, b/c © concerns drive far more than whether one of the exclusive rights is violated. Incentives to create works of original expression in the first place, which is driven by © incentives [for fridges, sure]. Access can occur w/o infringement and that is a core © concern and we shouldn’t treat that narrowly so that we don’t think that only infringement raises core © concerns.
Q: goes to Slover’s point: is there a way to reform 1201 to include things like circumventing TPM to watch movie for free, but excluding garage door openers? One idea: permanent exemption for software essential to operation of a device. Would that strike a proper balance?
Zuck: Discussion on embedded software roundtable was quite robust, but the same dynamism in product offerings still applies to “embedded software market” as well. Tivo: the hardware was given away as a loss leader, and the embedded software and services associated w/it created the value. [You could do that w/cars too—but usually you do that with leases, and there isn’t an obvious case to be made that you should be able to do that by preventing replacement parts from being used.] Couldn’t support broad exclusion of embedded software.
Panjwani: a permanent exemption for embedded software would be a good start, addressing overhang of issues like 3D printers. There are also a number of issues around traditional expressive works. [As indeed software must be to be protected by ©.] Exemptions involve appropriate balance b/t access for fair use purposes; 1201 also substantially burdens those uses. Caution that any attempt to create a permanent exemption would instead turn into a fight about what a software-enabled device is; could spend years hashing out a definition. End result would be fighting not about © infringement but about the device definition.
Pierre-Louis: tread carefully w/blanket exemptions. Software is expressive: we use software to protect software. Blanket rule could swallow entire industry. Games are played on game consoles; TV; movies; all manner of distribution b/c consumers demand it. So we have to tread carefully b/c implicates more than just a tractor.
Band: Switching hats to Owners’ Rights Initiative: we could support that, though a nexus requirement would be better. Embedded software exemption would be a good start. There could be exceptions to the exception for game software. Bigger point: Congress in 1201, and I was part of those discussions, wasn’t thinking about tractors. Fact that we’re talking about tractors does suggest a serious problem. Internet of things: CO as sudden regulator of the entire manufactured world. We went from 2 exemptions to 22; in the next cycle it will continue to grow geometrically.
Lowe: Emphasize parts on a car that used to be repairable using mechanical means now require software access. The software is equivalent to the parts; taking away the right to repair threatens our industry. People should be able to modify their cars in their garages; your suggestion is a good one. Mechanical functions taken over by software, including windshield wipers: the noncopyrightable function should not be protected.
Dow: Embedded software “necessary to make a device run”—I’m not sure how to interpret that, which highlights difficulty of drafting. Thinking back to early cases, RealNetworks—authentication sequence; you needed to authenticate your device to ensure you were talking to a real server. If you spoof the authentication sequence, that was really software, and a lot of what goes on in TPMs is authentication. Urge caution of impact on totally different contexts. [Which is exactly what the car folks are saying!]
Q: devices w/ and w/o expressive content? One testimony said that the software was the same no matter what the device does. Can we distinguish based on what the device does?
Pierre-Louis: we have consoles; we have servers in the cloud. We have different kinds of interoperability/authentication. It takes a lot of thinking about how those work before we get into rules. Software that operates the machinery might be copyrightable, but I don’t know enough about the tech to say it shouldn’t count. Some of these other areas, they are probably doing copyrightable things, and it’s not for me to judge. There are rationales behind each of our uses of software, and there are unintended consequences. For us, the leaks are the business; once they’re out, they’re out.
Zuck: There’s certainly software itself; we are interested in using these protections. A lot of different licensing models for software itself, and protecting dynamism is really important. Allows different models of making software available to different sectors. I’d rather approach this from the other end. 1201 process in place: maybe it’s better to try to streamline renewals, once you’ve identified a clearly acceptable use. Seems to be a working process [heh]. Rather than trying to go through and predict legislatively into the future. [Because the 1201 prediction was so bad … no, wait, that’s not your position. Everyone likes predictions that predict that they should continue to make money.] The current system is working. I don’t mind the default answer being no. The notion that people are afraid to start new businesses b/c of © liability is not a problem—it’s not that confusing what is and isn’t legal. [Uh, ok.]
Greene: We spent a while on marketplace implications of effectiveness/ineffectiveness of TPMs, but not on public safety. Consider chilling effect on security research. Geometric increase in connected devices; 1201 is a tremendous obstacle to security research to ensure that cars, airplanes, fridges, or TV sets are in fact secure. When we set access as the threshold and tell them that 1201 is meant to stop that, as opposed to protecting against infringement, then we’re telling the public that market interests are more important than public safety and health.
Adler: w/r/t those concerns, they’ve been addressed to some extent by Congress in exemptions in the statute, and every 3 years we revisit many of those issues in rulemaking. Goes back to the Q of whether the issue of infringement becomes the tail that wags the dog of 1201. 1201 was a recognition that wholly apart from infringement there was value to access that had to be w/in rightsholder to control, otherwise they wouldn’t provide it online. Courts haven’t had any difficulty dismissing the notion that fair use contains a right of access to a copyrighted work. It never has. [Our folks have lawful, paid-for access; what they need is the ability to make creative uses. This is the problem of the merged access/rights control.]
Panjwani: the discussion we’re having now about embedded software highlights the danger of setting defaults as no. The growth of these other issues shows that what we thought was small actually implicates a vast range of economic activity. Reevaluate whether the default being no still makes sense! The examples of successes—Chamberlain & Lexmark—exist, but MDY v. Blizzard and RealDVD are counterexamples rejecting the rationales of those cases. We never get to core © issues b/c of 1201. Adler says access is a right: but that’s not what Congress said—they said they were creating an ancillary right for the purpose of discouraging infringement of 106, due to easy reproducibility of digital content; not a standalone right.
Zuck: Public safety: it’s equally important to recognize that there are public safety implications to tinkering w/embedded software. Medical devices. [The CO is clearly the correct decisionmaker here, amirite?]
Adler: for my industry, library elending or print on demand or rental on demand, the issue is software being used to provide access to the expressive work that’s protected by ©. Don’t let difficulties created by dual ID of software as expressive work and uses to control access to a separate work to place those works in a position Congress didn’t intend. Right of access wasn’t considered only w/r/t piracy—it was specifically that Congress sought to create online markets for © works. That’s why access had to be controlled v. what you intended to do with the work once you had access. [What would the Framers have thought about cell phone unlocking and digital tractors? Originalism is often unhelpful.]
Band: there’s a big difference b/t access to something you paid for and access to something you didn’t pay for. Congress did create some permanent exceptions, but there wasn’t one for the testing of the VW for purposes of determining that they were committing an enormous fraud on consumers around the world. No one is stealing the VW to do the testing; they’re buying the VW and trying to figure out how the software works, but much of that isn’t covered by the exemptions that exist. Yes, there are people who take the risks, but many of them have decided it’s not worth it. If you’re at an institution and you know there are colorable legal issues and you have funders to please, people decide not to go there—it’s hard enough to get tenure without that. Consumer protection issues, safety issues, many other things not covered and to hope that rulemaking will continue to work for you every 3 years is insufficient; we could make relatively simple narrowing changes to the DMCA.
Greene: Reiterate chilling effect. Public safety concerns: I wasn’t suggesting that’s an equity for the CO to weigh; in fact, the inquiry should be limited to whether the proposed use would constitute an infringement. Making the point that by expanding the inquiry beyond infringement implicates significant public policy concerns. The idea of access as a threshold—the industry is trying to have its cake and eat it too. Other industries can’t protect their products w/© for the underlying work. At some point 46% of Americans engaged in piracy; clearly it’s not working, yet the vast majority ended up buying more digital products than those who didn’t. So we should rethink ithe value of limiting access; it doesn’t seem to be doing what it was supposed to do.
Q: how prevalent is it for 1201 to be used for what you might regard as anticompetitive purposes or to enforce 1201 against consumer products? We’ve heard about Chamberlain and Lexmark, but those are older and Ps lost, but we’ve heard about chilling effect.
Dow: What I see is largely anecdotal and in context of rulemaking. 27 sought and 22 granted, so the process is working. To Panjwani: not correct that access was not separate from infringement. House report said: TPMs can be deployed not only to prevent piracy but also to support new ways of disseminating and safeguarding legitimate uses. Purpose: incentivize new business models. Those are the business models we’re engaged in. We have apps allowing you access to our TV channels, watch on the go. Access is controlled through authentication. [And you wouldn’t sue someone for © infringement for unauthorized use?]
Panjwani: We assume that many people are law-abiding; if the activity would be prevented by 1201, then much of the activity wouldn’t be out there. We filed to clarify the legal ruling on 3D printers, which led to a voluminous opposition by Stratsys. Another competitor just announced end of life of one of its printers, which had lock-in to proprietary filament, showing need for exemptions.
Pierre-Louis: 1201 succeeded in getting works online. [The same way I succeeded in getting rain to fall the past week.] Businesses like certainty. Invest more when they know they can get a return. That’s been proven. [Except if you attended the content industries’ presentations at the 512 hearings.]
Weissenberg: Tracfone filing: has used the DMCA to enforce its anticompetitive lock-in many times.
Band: Trump thinks unpredictability is a virtue, but it’s not good for clients. After DMCA passed, plain reading of 1201(a) showed it could be used anticompetitively. Had to counsel clients that way. Chamberlain allowed a change, but then Blizzard v. MDY came by and rejected Chamberlain—now has to tell clients they can’t compete in the aftermarket if they’re in the 9th Circuit. Look at who is opposing exemptions to see its anticompetitive use: market leaders in phones, 3D printers, auto manufacturers.