By Eric Goldman
Sen. Wyden and Rep. Issa have released a draft of OPEN: Online Protection & ENforcement of Digital Trade Act, intended as an alternative to SOPA/PROTECT-IP. See my prior posts opposing SOPA and linkwrapping the discussion. Unlike SOPA’s disgustingly blatant rent-seeking, which was such an over-the-top abuse of the legislative process that it did not (and could not) support a principled or even intelligent conversations about it, OPEN provides a useful starting point for a sensible conversation that could actually lead to acceptable compromises. For that reason alone, I think Congress should immediately stop all work on SOPA/PROTECT-IP and redirect that energy towards vetting this proposal. Having said that, for reasons I’ll explain in a moment, I continue to believe the assumptions underlying SOPA/PROTECT-IP and OPEN are misguided, meaning that forging a compromise from OPEN’s more sensible proposal may be tricky.
Before I get further into substance, two process notes:
First, SOPA was the product of rent-seekers who were talking only amongst themselves and legislators tethered to their campaign contributions. The drafting process was disturbingly closed-door and exclusionary, exactly the kind we wish didn’t take place in our representative democracy. In contrast, the OPEN sponsors want to have a dialogue about their ideas. In support of that, they have posted the draft to a website that allows comments and discussion. This is the way our democracy SHOULD work. Why is such an open process the exception instead of the rule?
Second, OPEN is a comparatively svelte 18 pages focused mostly on one core concept, compared to SOPA’s 78 page monstrosity that advanced about a dozen different substantive proposals. I can’t tell you the number of times I’ve seen very smart people stymied to keep all of SOPA’s moving parts separate, and the failure to do so meant that they were conflating different parts of the statute in ways that prevented productive discussion. (Just two examples: the Colbert Report, where Zittrain mostly focused on SOPA’s felony streaming provision while his counterpart was mostly talking about the cutoff provisions; and Business Insider’s infographic where the felony streaming sanction was presented as a remedy to the cutoff provisions.) By reducing the number of topics at issue, OPEN substantially reduces the chance that policy discussants will simply talk past each other.
The law contemplates that rightsowners can file a petition against rogue websites at the ITC, an independent federal agency best known for its adjudication of certain patent disputes. In response to the rightsowner’s petition, the ITC will conduct an administrative adjudication. If the ITC determines that the website is a rogue website, then (1) the website is required to cease its conduct (not sure how enforceable that is), (2) the site also will be subject to any other unspecified consequences following from its determination as a rogue actor, and (3) most importantly, the rightsowner can take the ITC determination to payment service providers (PSPs) and ad networks and have them cut off the flow of money to the rogue website. The PSPs and ad networks would be protected by several immunities for trying to comply with the orders or their other efforts to protect the public.
This makes OPEN similar to SOPA in that it seeks to cut off funds flowing to rogue actors. However, among other key differences, PSPs and ad networks have no legal obligations until the ITC makes a ruling. In contrast, SOPA imposed cutoff obligations on PSPs and ad networks based merely on rightsowners’ unsubstantiated assertions.
Substantively, some of the things I liked about OPEN:
* it situates the discussion about “rogue websites” in foreign trade policy. This fixes SOPA’s overinclusive application to both domestic and foreign actors. However, if we really think rogue websites are a transborder enforcement problem, there are many other trade policy solutions that might be better options to consider—the most obvious being transborder enforcement coordination like the FTC does with its foreign counterparts.
* OPEN doesn’t touch the domain name system or search engines. SOPA had the potential to destroy the DNS and to jeopardize search engine functioning. OPEN sidesteps both pitfalls.
* OPEN builds in some due process before any formal legal obligations attach. As we’ve recently seen, due process is actually quite important, and we suffer from its absence. I say “some” due process because I’m not sure how much due process will attach in practice. For example, I have some concerns about the notice provision–not every targeted website will receive notice of the ITC investigation. However, I did like that any website the ITC labels as rogue can correct any identified problems, reapproach the ITC and ask it to remove the “rogue” determination.
* the definition of rogue website is tightened up substantially. It requires three elements:
a) a “non-domestic domain name,” which requires that the registry, registrar and registrant all have to be located outside the US (I’m not sure what “located” means in this context). Venkat asked me what happens to a .com registered with a foreign registrar; I believe OPEN does not apply to this domain name.
b) conducting business in the US; and
c) “has only limited purpose or use other than engaging in infringing activity and whose owner or operator primarily uses the site to willfully engage in infringing activity.”
The last element, in particular, is quite restrictive by requiring willful infringement. The meaning of the word “willful” is notoriously murky (see, e.g., the multitudinous Supreme Court cases over the word), so the statute would be improved by using a more detailed synonym. No matter what, though, willful is a high scienter level that should easily exclude most legitimate players. The statute further expressly excludes any sites that:
– follow good notice-and-takedown procedures
– qualify for 17 USC 512 (the DMCA online safe harbors) [this means that the statute sits next to 512 instead of rendering 512 moot like SOPA threatened to do], or
– distribute “copies that were made without infringing a copyright or trademark.” I’m not 100% sure what this means. It apparently excludes websites reselling goods covered by the First Sale doctrine. I presume that the exclusion includes sites that sell legitimate knock-off goods, such as replicas of goods that aren’t protected by copyrights or trademarks.
* if a PSP or ad network fails to comply with an ITC order, the only consequence is that the DOJ can seek injunctive relief. Rightsowners do not have a private cause of action in those cases. As discussed below, this doesn’t eliminate all PSP/ad network exposure to rightsowners, but rightsowners can’t introduce evidence of ITC orders in any civil suits they bring against PSPs or ad networks.
* on the trademark side, it expressly limits its applicability to counterfeiting (although there is an erroneous cross-reference in the draft). Presumably, dilution or garden-variety trademark infringement disputes don’t qualify under the statute.
What’s Not Good
Substantively, some of the things I don’t like about OPEN:
* OPEN still contemplates reestablishing a Fortress USA. Fortress USA marginally makes sense regarding the shipment of physical goods across geographic borders. It makes zero sense for digital bits zinging around the borderless network.
* in particular, because OPEN would burden only US-governed PSPs and ad networks, it may drive websites—including legitimate websites who want to reduce their risk of being mistargeted—to shift their business to foreign-based PSPs and ad networks. If lots of businesses make a switch based on these concerns, OPEN could counterproductively result in net financial losses for the US economy.
* similarly, foreign websites can opt-out entirely of the ITC process by consenting to US judicial jurisdiction. I like the idea of an opt-out, but imagine if other countries offered the same quid-pro-quo of allowing US websites to opt-out of some nasty foreign process so long as the websites consent to jurisdiction in their countries. I think we’d be outraged and insulted; which is how I would expect foreign countries to view this quid-pro-quo. Cf. Venkat’s recent post on Facebook v. Faceporn. Then again, other countries might think it’s a pretty good idea, leading to a proliferation of transborder quid-pro-quo jurisdictional offers.
* designating the ITC to conduct the investigations is a little odd. First, the ITC is an administrative agency, not a federal court. I don’t fully understand all of the implications of administrative vs. judicial review, but I believe there are substantial procedural differences that could lead to important substantive differences. Second, the ITC has been gamed in the patent world (see, e.g., my colleague Colleen Chien’s research on the ITC explaining how the ITC hears many US company vs. US company disputes), so I fear similar gaming will emerge. For example, a rightsowner chasing a rogue website could simultaneously pursue a domestic court action, a foreign court action and an ITC proceeding. How would these types of parallel proceedings play out in practice? We’re still trying to resolve the parallel proceeding problems in patents.
* like SOPA, the bill covers copyright infringement, trademark infringement *and* 1201 circumvention. I don’t understand why the circumvention issue is getting equal billing or how often transborder circumventions are a real problem. Seeing how 1201 circumvention lawsuits have devolved into anti-competitive enforcements, picking up the circumvention piece could increase the risk of competitive misuse of the statute.
* like SOPA, the definitions are vague. Consider, for example, the definition of Internet advertising service:
The term Internet advertising service means a service that serves an online advertisement in viewable form for any period of time on an Internet site.
Hmm…what does that mean? Notice that the definition doesn’t directly distinguish between third-party ad networks and sites that sell their own ads. I think in practice sites that sell their own ads drop out of the statute, so one possible implication is that more sites will ramp up their own ad sales. (This is doubtful, but just throwing the possibility out there.) I think the focus on “viewable” is interesting; are audio-only ads excluded? And what does it mean to “serve” content? This contemplates a specific technological interaction that I don’t fully understand today and will almost certainly evolve over time.
Why I’m Not Enthusiastic About OPEN
Even though OPEN is worth discussing intelligently, unlike SOPA, I believe it’s based on two underlying assumptions that aren’t fixable.
First, like SOPA, OPEN assumes there is a problem with foreign rogue websites that needs to be solved. I’m not saying there isn’t, but the policy discussions have been startlingly devoid of reliable and credible facts demonstrating the nature and scope of the problem.
Instead, the evidence in support of a rogue website “problem” typically consists of two main threads: (a) people are dying from counterfeit drugs, and (b) bad guys are “stealing” our stuff. With respect to the former, I’ve never seen anything more than ad hoc assertion; but if there’s a real problem, counterfeit drugs can be fixed with a highly targeted solution. With respect to the latter, it’s hard to give those arguments much credit. After all, all rightsowners’ arguments are inherently self-interested: it’s in their financial interest to say that they would like to make more money than they are making. It’s also in their interest to bemoan broad sectoral changes in the economy as evidence that someone is capturing money they think they are entitled to (and to use rent-seeking to thwart those broad sectoral changes). More importantly, there is lots of evidence that a lot of rightsowners are making a lot of money today, both via the Internet and more generally. So it’s hard to break out the quantity of actual economic losses that rightsowners are truly suffering when those claims are intermingled with rightsowners’ general rent-seeking efforts.
Therefore, until the rightsowners offer us more than the trumped-up BS already-discredited statistics, I’m still not clear on the problem, how bad it is, how any legislative solution would remediate that problem, and if the collateral consequences of the effort to remediate the problem are greater or less than the problem itself. OPEN does nothing to fill the void of supporting foundational evidence of the problem, so it’s hard for me to be enthusiastic about its solution.
Second, and more importantly, attacking the money supply to supposed bad actors remains too blunt an instrument. I may be truly on my own on this point, as many people I respect–including, notably, Rep. Lofgren–are prepared to embrace the policy solution of cutting off money flows. However, by embracing an attack on the movement of money, OPEN replicates one of SOPA’s sins. If a player is engaged in legitimate and illegitimate activity and its money supply is cut off, both activities go down the tubes. In contrast, one of the positive aspects of 17 USC 512(c) and (d) is that they require the copyright owner to identify infringing items and target only those items. Giving rightsowners a remedy that would affect an entire site for only some items on the site goes too far.
The OPEN bill tries hard to minimize overbreadth by narrowly defining the targeted websites. Perhaps this definition is narrow enough that there won’t be much collateral damage. However, in practice, regulating money flows nevertheless could have pernicious effects in the field. A PSP or ad network drawn into an ITC proceeding frequently will “voluntarily” choose to toss the targeted website before the ITC proceeding reaches its conclusion—even if the ITC proceeding would have rejected the challenge. Furthermore, rightsowners still will send cutoff notices to PSPs/ad networks without filing any ITC petition, and the PSPs/ad networks will often honor them as a way of preempting an ITC proceeding.
What this teaches me (in combination with the Elsevier v. Chitika case) is that PSPs and ad networks need robust statutory immunities which are not based on a notice-and-takedown scheme. On the trademark side, the need for an immunity became clear after the sloppy language in Gucci v. Frontline. On the copyright side, 512 doesn’t cover PSPs and ad networks, probably because in a million years the safe harbor drafters never thought PSPs and ad networks would be liable for third party infringing activity in the first place. Now that we’ve seen copyright law and trademark law creep much further than we could have imagined in 1998, we should plug this liability hole completely. If OPEN proceeds, it should have a broad-based immunity for PSPs and ad networks with the idea that rightsowners are getting a specific remedy against them in the new law.
While OPEN can’t really be fixed to resolve my two structural concerns, my hope is that the discussion about OPEN will force rightsowners to provide *credible* evidence of harms that they or consumers are suffering (no more self-serving hype, please), and that such evidence will force us to think carefully about how “rifle shot” solutions (as opposed to shotgun solutions) can ameliorate those harms. If we have a discourse that even slightly resembles this ideal, then OPEN will be successful no matter what final outcome we reach.