Transparency Report Update: July – December 2020

It’s time once again for the bi-annual update to Automattic’s transparency report covering the period from July 1, 2020 – December 31, 2020. As ever, we’re providing details about government requests for user information as well as government demands for content removal. We’re also providing information on intellectual property disputes such as DMCA takedown notices and reports of trademark infringement.

Transparency remains a top priority for us and we’re always looking for better ways to present this valuable data. Last year, we announced that we were making a CSV file of our transparency report data available for the first time. We’re thrilled to continue this approach for the reporting period between July through December, 2020. For those looking to slice and dice our data more granularly, you can download a copy here.

For the first time, we’re also reporting on the number of requests we receive from users exercising their privacy rights. Specifically, these statistics include data access requests, deletion requests, and do not sell requests. You can learn more about these requests and Automattic’s commitment to user privacy here.

Further, we previously mentioned that we’ve been working with our colleagues at Tumblr to better align our data gathering processes and present a unified transparency report under one roof. We’re now one step closer to this goal and you can find our new landing page here which will direct you to the specific report you’re looking for.

Below, you can find a few more interesting highlights!

Intellectual Property

In our last update, we mentioned that the total volume of DMCA takedown notices received in the first half of 2020 was just shy of 10,000—nearly double the volume of the previous six months. We also pointed out that this spike was the result of a single complainant: the anti-piracy company, 3ants. The Madrid-based firm was our top complainant by volume last reporting period and they hold that title again for notices received between July through December, 2020. However, while many brand protection companies have little interest in open dialogue with hosting platforms, we’ve been really pleased by how receptive 3ants has been to our feedback.

Aside from sheer volume, the primary challenge with processing takedown notices from companies that submit them en masse is that the notices typically require additional scrutiny by our Trust & Safety teams to ensure validity. It’s common for anti-piracy firms to take a “kitchen sink” approach with their notices which often cite content that we don’t host or that we have already removed, and they tend to bury the required statutory DMCA language in a sea of unnecessary legalese. Perhaps most frustrating is that we rarely receive any replies from these types of companies when we ask for clarification about a notice.

So, given the landscape you can imagine our surprise when we contacted 3ants with our concerns earlier in the year and received a prompt reply! In addition to welcoming our input, 3ants committed to adjusting their approach for takedown notice submissions to Automattic. Since then, their notices have arrived through our DMCA form, each notice only cites content from a single website rather than dozens, and we now receive replies to our requests for clarification. Each of these improvements means we can more quickly review and process takedown notices from 3ants which is a big win for both rights holders and site owners alike.

Unfortunately not every complainant is as cooperative as 3ants. For years we’ve been speaking out against abuses of the DMCA such as the use of automated systems which flood platforms with takedown notices regardless of context. These methods are often prone to error and make it difficult for platforms to prioritize valid notices submitted by individual rights holders.

But what happens when human-generated notices cause similar issues? This is the situation we found ourselves in during the second half of 2020. Across those six months, we received over 2,800 takedown notices which—despite being submitted by humans—contained many of the same issues we encounter with automated notices. We informed the complainant of each instance that they targeted content hosted elsewhere only to receive multiple follow ups demanding that we disable access to the materials (among other threats). These follow ups, which fall outside of established DMCA processes, are a major time suck for platforms who are verifying whether or not new and valid notices are included in these replies.

This type of behavior undermines the spirit of the DMCA and is ultimately detrimental to rights holders. While platforms like ours are forced to engage with these tactics, individual content creators must wait patiently as they navigate the DMCA process in a more typical fashion. Lastly, it also obscures transparency report data as it’s challenging for hosting providers to effectively log excessive follow ups which fall outside of expected DMCA processes.

Government Takedown Demands

Fortunately, both IRU requests and government takedown demands have remained relatively low, though we did see slight upticks in both categories compared to our previous report. IRU requests increased from a total of 3 in January – June 2020 to a total of 11 in July – December. Similarly, the volume of government takedown demands accompanied by a court order jumped up from 12 to 26 over the same timeframe.

The vast majority of these court orders came from Turkey with a total of 23 over the last six months of the year. Turkey also holds onto their top position as the region where the most WordPress.com sites are geoblocked with a total of 470 blocked sites (followed by Russia at 343 and Pakistan at 301). 

Although demands from Pakistan have decreased following their meteoric rise to 3rd place in terms of geoblocked content over the past several years, their demands still rely on some questionable reasoning. For example, we recently received a content removal demand aimed at a blog post which “harms the reputation of government officials.”

We hope you find this information interesting and helpful. As always, please contact us if you have any questions or suggestions about these reports.

Automattic at RightsCon 2018

On May 16, 2018, Automattic is teaming up with Global Partners Digital for a panel on content moderation and changes to legal liability for online platforms being considered by governments around the world.

During our RightsCon session, Content Regulation in the Digital Age, we will explore the role of publishing platforms and social networks in the face of increasing calls to regulate content, and consider the definition of ‘content’ as it relates to ever expanding forms of expression facilitated by the internet. Automattic has long been a strong advocate for our users and for transparency in reporting the actions we take with regards to content moderation on our platforms. We look forward to sharing details of this work as part of the discussion at RightsCon.

Moderated by Charles Bradley, Executive Director of Global Partners Digital, the hour should prove to be a lively discussion between panelists as well as an engaging Q&A with conference attendees. During the week, Global Partners Digital will also be launching their white paper on a rights-respecting model of online content regulation.

This is Automattic’s second consecutive year attending RightsCon. We’re pleased to return and connect with industry professionals and digital rights groups from across the globe. Our trust and safety team will be present throughout the week. Also keep an eye on the schedule for details about our official after party following the conference sessions on Thursday, May 17.

RightsCon Toronto takes place Wednesday, May 16 to Friday, May 18 at the Beanfield Centre at Exhibition Place. Tickets are available now!

Leave No Trace!

Earlier this year, we talked about the dangers of relying on third-party bots to chase down potential copyright infringement, and the ironic circumstances that ensue when the rights holder’s own content becomes a target. It’s a frustrating situation for everyone, but this kind of oversight is no surprise when automation is let off the leash. What’s even more shocking are cases in which this paradoxical level of enforcement is employed by humans, as we recently experienced with Subaru for America.

The car manufacturer is currently sponsoring a nonprofit organization, The Leave No Trace Center for Outdoor Ethics. For the uninitiated, “Leave No Trace” is the concept of enjoying the wilderness responsibly, with an eye towards conservation and preservation of natural beauty. A noble cause to be sure, but two WordPress.com users discovered the hard way that a corporation’s actions can sometimes speak louder than its words.

In the application for The Subaru Leave No Trace Traveling Trainer Program, it’s suggested that, among other things, applicants should be able to generate social media content via creating blogs, posting photos, and creating videos. This is exactly what Sam and Jenna attempted to do when they created this blog as part of their application process.

Surprisingly, however, Subaru was less than thrilled with this harmless use of the brand name, and accused our users of trademark infringement and cybersquatting. Apparently the legal department at Subaru wasn’t aware of their company’s sponsored initiative. Yikes!

We passed along the notice to Sam and Jenna with the understanding that we would not be taking any action against their blog as their content does not constitute trademark infringement. In good faith, though, our bloggers took the high road and removed most instances of ‘Subaru’ from their blog to appease the automaker’s demands.

While Subaru reminded us that it “aggressively protects its trademarks and other intangible assets,” it’s unfortunate in this case that they didn’t maintain that same level of overzealous enthusiasm for our users or the great outdoors.