The data contained in the Atlas of Surveillance is open-source intelligence, or OSINT. This is a term used to describe gathering information that already exists online—from news stories, social media posts, press releases, or documents buried in government websites, often turned up through using advanced search engine techniques. Our data collection took two different forms:
Building off earlier EFF crowdsourcing projects, we developed a tool called Report Back to automatically distribute small assignments to students and volunteers.
When a user visits Report Back, they are assigned a small research task consisting of a particular location and a technology (e.g. body-worn cameras and the Tulsa Police Department). The researcher then spends 20-30 minutes looking for news articles, press releases, meeting minutes or other online documentation, and logs their research in our database. A large number of these assignments are based on leads generated by GovSpend’s database of government procurement records.
As of July 2020, more than 500 students and volunteers have contributed research to the project. Each line of data was then double-checked by multiple interns from the University of Nevada, Reno and EFF staff.
EFF is not the first organization to build databases of surveillance technology. Most of these previous efforts, however, have either focused on one particular technology across the United States or all technologies in one particular place. For the Atlas of Surveillance, we began collecting these public datasets from journalists, other non-profits, government entities, and sometimes surveillance vendors themselves. We then converted this data to match the Atlas of Surveillance format.
Open-source intelligence and crowd-sourcing come with their limitations. First, the information is only as good as the source: sometimes government agencies withhold information and sometimes journalists misinterpret information. It’s possible that while there is information about a technology being adopted, the technology was later abandoned, and no reporters wrote about it. With thousands of data points to go through, it is impossible to exhaustively fact-check each one, despite the multiple reviews by students and staff. In particular, documenting the use of face recognition has proven challenging because of the changing policy landscape that has resulted in local governments abruptly freezing or abolishing the use of biometric identification software.
The Atlas should not be interpreted as an inventory of every technology in use. It only represents what our team documented after a year and a half of research.
In short, the Atlas of Surveillance serves as a resource, a tool, and just one way to understand the growth of surveillance technology in our communities. We hope that it will be the starting point for many other research projects to come.
Updates and Corrections
Should you identify data that needs to be corrected or updated, please email firstname.lastname@example.org. We will be adding to and updating the database periodically.
Please note that points are placed on the map using an automated system and, like all automated systems, it makes mistakes. However, those mapping errors do not impact the text search.