Today we are happy to announce the Shapeways transparency report for 2016. This report is designed to give everyone in the Shapeways community insight into how our systems governing intellectual property disputes and third party access to Shapeways user information work.
What is a transparency report, and why publish it?
A transparency report is a public document that sheds light on how internal processes here at Shapeways work in practice. While the entire Shapeways community is impacted by our policies covering things like copyright disputes and privacy, in most cases individual disputes over those issues happen behind closed doors. This is a good thing in specific cases – community members should be able to resolve their differences outside of the spotlight. However, it can also make it hard for people who are not directly involved in a dispute to understand how the process works, or how those processes are working in aggregate.
The transparency report helps to summarize how our processes work and to give the entire community a better understanding of the trends emerging from them. It also helps the larger public and policymakers understand how systems grounded in law play out in reality. As we note in the report, it is impossible to evaluate the laws that control how Shapeways operates without understanding how those laws impact Shapeways and the Shapeways community. You can also compare the 2016 report to the 2015 report, which is available here.
What’s in this report?
I encourage you to check out the report itself, but three high level points are worth mentioning. The first is that intellectual property-related takedown requests almost doubled in 2016. This may not be a surprise since our community was larger in 2016 than it was in 2015. However, it is noteworthy.
The second is that 20% of trademark-related claims were withdrawn after low levels of scrutiny. As we have advocated for some time, there are real problems with current statutory safe harbors related to trademark infringement claims. As a result of these weaknesses, we evaluate most trademark takedown requests for facial reasonableness. In practice, this means verifying that the model targeted by a takedown requests has not been obviously misidentified (for example, if the trademark incorporates a common noun and the model just represents that common noun) or that the trademark is not being used in a way that merely connotes compatibility. It is something of a shock that 20% of all trademark claims fail what should be an exceedingly low bar for accuracy.
The third point of note is a positive one. In the 2015 report, we raised concerns around takedown requests that ambiguously claimed that a model infringed on a number of different types of IP. These types of requests — often as a form letter— made it hard for users to determine exactly what type of infringement was being alleged. One of the results of this ambiguity was to remove the disputes from the user protections in the Digital Millennium Copyright Act (DMCA) process. In 2016 we saw a dramatic reduction in the use of these sorts of combined notices (from 70% in 2015 to 9% in 2016). We appreciate that rightsholders are using more specificity in notices.
That’s the high-level summary of the report. The whole thing has more details, numbers, and even some graphs and charts. We encourage you to take a look and let us know if you have any thoughts in the comments below.