Lumen Researcher Interview Series: Ernesto van der Sar of TorrentFreak

Recently, the Lumen team had the opportunity to interview Ernesto van der Sar of TorrentFreak.

Ernesto and the TF team are frequent users of Lumen’s database for their journalism and research, often doing deeper dives into a particular notice or set of notices, so we were glad to be able to get into the details of how they have been able to use Lumen in their work.

This interview, which took place over email, has been edited for length and clarity.

Background Information:

Full name: Ernesto van der Sar

Organization and Position at the organization: TorrentFreak

Number of years at the organization: 16

Tell us some details about TorrentFreak , its history, and your work there

TorrentFreak is a niche news site that closely follows the file-sharing and piracy ecosystem. Over the years we have documented many high-profile copyright cases, piracy habits, legal developments, and various takedown trends. We work with a small team of just two writers and strive to highlight news that’s not (yet) covered by mainstream publications.

When we launched in 2005, most news reporting was limited to how rightsholders perceive matters. TorrentFreak aims to hear the ‘other side’ as well and to add – in our view - more balance to the discourse. This is in part done by allowing the operators of major pirate sites and services to share their views. In addition, we also highlight contrarian academic research and questioned results and conclusions from industry supported studies. Our goal is to cover news with a healthy balance, by highlighting arguments from all sides.

In our work we keep a close eye on takedown trends, including the many mistakes that are made. This is where Lumen comes in. The project is a crucial resource for us which has resulted in well over a hundred articles over the years.

How did you or TorrentFreak first become aware of Lumen? When did you start using it for research?

The first time I came across Lumen was in 2007, when it was still called “Chilling Effects.” The operator of SumoTorrent, a popular torrent site at the time, noticed a reference to the database at the bottom of Google’s search results. This is the first time that Google’s DMCA removal activities came on my radar, years before it launched its own transparency report. At the time, the Google search engine had only received 177 takedown notices in the entire year (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2411915), so it was quite unique as well.

[Lumen note: Google’s transparency report states that it has, as of August 2022, received takedown requests asking for the removal of almost six billion URLs]

At which stage in your writing or work generally does Lumen usually feature? Do TF stories ever start with Lumen?

Lumen can definitely be a starting point for research. I regularly browse through Lumen, using various search terms to find potentially interesting notices. In other cases I start with Google’s transparency report, which then refers to Lumen for the finer details.

How do you use Lumen? How often do you use it in your research/work?

Every week I visit Lumen multiple times, often without a concrete goal. [Lumen: We love this!]

Just sifting through the notices is often good enough to stumble upon interesting leads. There don’t always result in news articles, but it helps to identify problem areas and stay on top of new trends. The searches are mostly manual and involve keywords, domains, senders, or rightsholders of interest. I also narrow down specific topics such as “court orders” and “anticircumvention” to search more narrowly.

How does the availability of takedown notices within Lumen enhance your ability to write on the topics you cover?

Lumen’s information is crucial to report on the takedown ecosystem. Without it, it simply wouldn’t be possible. It would be great if more services start sharing this information in the future.

What are some other resources or tools that actively feature in your work and research?

When it comes to reporting on DMCA notices we also use Google’s transparency report, which is very helpful. The same is true for GitHub (which is also in Lumen’s database now). Other transparency reports (Wordpress, Bing, Cloudflare etc) can provide interesting information as well but these are more limited when it comes to reporting the contents of specific takedown requests.

What additional tools/ features / UX and UI would enhance your ability to use Lumen?

It would be very helpful to be able use the search function to filter based on domains.

As a journalist working on copyright and tech law, how important, in your opinion, is transparency (on the part of OSPs or otherwise) with respect to takedown notices sent and received.

Extremely important. We have highlighted dozens of serious mistakes and errors over the years. Many of these resulted in the removal of content without a proper reason. When these matters are highlighted in the press the senders often take steps to correct the mistakes and prevent them from happening again. This would not have happened without the transparency that Lumen, along with some OSPs, offer.

” What do you think of Lumen’s draft "Statements of Best Practices for Takedown Transparency," developed during its recent 2022 research Sprint?

Great initiative which covered many important angles already. Personally, I’m quite frustrated by the high number of imposters. This is a major problem in my opinion that should be addressed as it pollutes the current data.

[Lumen note: several members of the Lumen team are working on this topic right now!]

Where / how would you like to see those statements developed?

It might be a good idea to have a list of X key basic data points for every transparency report. That would be the bare minimum for what a decent report entails. Then there can be a much more detailed document that outlines all suggestions. I’m happy to share some thoughts. It’s probably a good idea to hear all important stakeholders, including senders, recipients, researchers, and the broader public.

Do you have any recommendations for academics/journalists/activists working with Lumen data, or on takedowns generally?

Not really, just keep digging.

What, in your opinion are the biggest “holes” in global takedown data? If you could have any one chunk of data you could, what would it be?

More detailed notices from more OSPS. YouTube is probably on top of my list. Notices that are sent to ISPs should be public as well. There are millions of these every year, but there is no accurate information available on the exact volume and how it changes over time.