A response to Senator Tillis’s questions regarding his proposed draft legislation and possible changes to Title 17.
Dear Mr. Watts, Mr. Greenberg and Senator Tillis,
My name is Adam Holland, and I am a Project Manager at Harvard University’s Berkman Klein Center for Internet & Society, where I oversee the day-to-day operations of the Lumen database project . I write to you today in my individual capacity to answer some of the questions you have posed to stakeholders regarding your DMCA Reform Bill, and to suggest a few additional changes.
Lumen is an independent research project that studies, and facilitates the study of the landscape for online content. This includes requests — based on legal or extra-legal theories — to remove materials (or links to materials) created or uploaded by Internet users. Lumen operates a research platform that invites rightsholders, Online Service Providers (OSPs), search engines, and other online intermediaries, as well as members of the general public, to share takedown requests that they send and receive concerning online content. Lumen maintains a database containing approximately 15 million such notices, all of which have been voluntarily shared with the project by their recipients and senders. Lumen makes those notices available to scholars, journalists, and others for purposes of research and analysis.
Lumen was formed as the Chilling Effects Clearinghouse in or around 2001, in the wake of the United States’ 1998 implementation of the Digital Millennium Copyright Act and its safe harbor provisions. The project was founded in recognition of the fact that the DMCA safe harbor in the US would change the landscape concerning availability of online content by incentivizing platforms to remove content in response to copyright claims in order to qualify for the safe harbor, and that it would do so through a process that took place largely out of the public eye. Against that backdrop, Lumen’s goals are to facilitate research about different kinds of complaints and requests for removal — both legitimate and questionable — that are sent to online publishers and service providers; and to provide as much transparency as possible about such notices in terms of who sends them, why, and to what effect.
However, no existing law or regulation mandates that removal requests or demands be made public or even available for study. Thus, all of the notices in the Lumen database have been shared voluntarily with Lumen. And, those companies that have chosen to share notices with Lumen decide which types of notice, and which data fields within a notice, to share. The set of data to which Lumen has access is unfortunately therefore limited, and of commensurately limited utility with respect to drawing conclusions about the efficacy of the DMCA. It is with that backdrop in mind, and with the insight that my experiences managing Lumen have given me, that I write with a response specifically to your question 7, regarding the shortcomings of the notice-and counter-notice sending process as well as its potential standardization.
If Lumen had a motto, it would be “good policies are founded on good data.” The best way to draft well-functioning new legislation and policies regarding online takedowns is to carefully study the takedowns that have already occurred and that continue to occur. Only by fully understanding that complex ecosystem and its stakeholders, the successes and failure of existing policy, can legislators move forward in an effective way, one that will not create new problems as it attempts to solve the old.
I therefore propose, related to Senator Tillis’s suggestion for “regulations for the communications that OSPs must deliver to a user when their content is taken down or had access disabled” that in addition to standardization of forms, which I would support, that in contemplating changes to Section 512, drafters should consider ways to effectively incentivize those who send and receive notices to disclose those notices in ways that allow for their ongoing study.
Of course the devil will be in the details, but as just one example, a provision could be added to section 512 mandating that recipients of DMCA notices and any related counternotices disclose the contents of such notices, along with any action taken, to an independent third party entity. It seems less germane to have DMCA senders disclose the notices they send, but that could also be a topic for consideration.
If it received adequate funding to handle the substantial new volume, these mandatory disclosures could easily be made to Lumen, the largest current aggregator of DMCA requests, which has twenty years of history managing and facilitating research within a database of such requests. Or, a new clearinghouse entity could be created, perhaps under the auspices of the Copyright Office. In the alternative, or as an option for smaller scale DMCA recipients, there could instead be a requirement for a standardized general disclosure to the public, in the form of a Transparency Report published regularly by each recipient of DMCA notices, at least any with a registered DMCA agent. There is already ample precedent for such reports, and many larger companies publish them.
These Transparency Reports would be required to take a standardized form, allowing external researchers, including Lumen, to easily collect, aggregate, and compare their data for further study. Access to the “raw” data could be limited to vetted and accredited researchers, while the public could have access to a version with private and personally identifying information (PII) masked, as question 7 suggests. Lumen also has extensive experience with handling PII within DMCA notices and would be glad to consult on this topic. Any necessary redaction should be undertaken by the least-cost avoider.
Were this level of transparency regarding DMCA notices and counternotices, as well as the ultimate actions taken on them by recipient OSPs to become standard, I believe that the resulting wealth of information and insights gleaned from it would make it much easier to answer all of the other questions Senator Tillis has asked, whether initially or as part of ongoing and regular evaluation of existing laws and regulations.
Draft language on this topic might look something like:
(X) In order to facilitate ongoing research and policy analysis, any online service provider, as defined in [Section NNN], shall, upon receipt of:
(1) a notice per the requirements of 17 USC 512(c) (3),
(2)a counternotice sent per the requirements of 17 USC 512(g)(3),
(3) a subpoena sent per the requirements of 17 USC 512(h), or
(4) a cease & desist letter sent regarding Section 1201 violation
transmit a copy of the notice, counternotice, subpoena or letter, formatted according to the criteria in[Section NNN], to the Lumen Database / Copyright Office’s clearinghouse for such documents described in Section (NNN);
or in the alternative, make a copy of the correctly formatted notice, counternotice, subpoena or letter available in a quarterly Transparency Report, as described in [Section NNN].
Project Manager — Lumen
Berkman Klein Center for Internet & Society
 The Lumen database is accessible at https://www.lumendatabase.org.