A project to lay bare what our consent allows
Whenever the Columbia SIPA event on Digital Identity, Security and Governance rolls around every year, I am in the audience. I do it for a few reasons. The participants are from varied backgrounds and leaders in their fields, the questions from the audience are incisive and challenging and I am very interested in learning about the policy context. It helps that the conference is in my backyard and it is free.
Of the situations that technology has created for society, one of the biggest is the rise of surveillance capitalism. Unfettered by geographic boundaries and controls and our continuous interactions with the devices that we have invited into our homes and our persons; most of our behavior is being captured and monetized by commercial behemoths. We were enticed by what appear to be free services. Our models are always based on prices, since we are used to the free market model, so what can be so bad about free? In addition, due to our evolutionary proclivities for visual engagement and gregariousness we flock to virtual communities, in search of dopamine hits for “likes”, “comments” and “views”. This situation could not have been foretold even by the very forces that are sucking up personal data and meta-data from all available sources.
Lately, more of us have come to realize the predicament that we are in. What can be done now?
The excellent Mapping Data Flows (MDF) Project has an answer. Visualizing the architecture of the data flows as stated by the companies themselves can expose the extent and intent of data collection. This establishes a cornerstone for a serious, informed conversation, leading to concrete actions by legislators or regulators.
MDF started by analyzing the commercial policies expressed in the Terms Of Service (TOS) by Google, Amazon, Apple and Facebook. MDF chose these firms because of the capabilities and reach of these enterprises, especially in the US. These are the same terms of service or end user license agreements that we blithely click through to get to the free services we crave. These documents cannot be read through thoroughly and understood by anyone, even experts, without expending a large amount of time and energy. The MDF team turned the documents that they read into data, and then created a visualization of this data on the Mapping Data Flows website. The website is interactive.
The unfiltered visual has three columns that consist of buckets. These buckets are sources of data, types of data and stated use. They had to do these groupings, otherwise the visualizations would have been unmanageable. In addition, this allows filtering by individual items in these buckets, to examine each of these in turn. The controls allow filtering by company and by preset groupings. In addition, the website allows comparisons between companies. There are some case studies that lead to further linkages from the TOS of a specific company to the data that is collected and used.
Their other intention is an attempt to start a conversation around data portability. Since the data collected by these companies are in silos, locked away and protected (hopefully) in company vaults. Unless they sell the data directly to other parties, which they do not, at least knowingly. Data portability will bring with it an aspect of ownership; although that itself is a loaded term. As this data can be replicated infinitely as well as stored away in the perfect memory of the industrial data management complexes of these enterprises.
You may have heard that the FCC has fined the cellular companies for selling cell tower and hence location data, this collection happens even if you turn off location services on your phone. You can see that this is one of the data sources on the MDF visualization. Maybe the only way to turn it off is to turn off your phone and take out the battery (not easy on an iPhone). Which means you can only make outgoing calls from a known location. The irony of this the fact that we pay the phone companies to put what is essentially an ankle bracelet tracking device on each of us. They happily record this data in their perfect memory and sell it to others. This was the subject of Carpenter vs. the US, a 4th Amendment case in the Supreme court about which I wrote an article.
The striking visual idea in the above diagram is the high density of the second column versus the first. This is because the data collected is used for many purposes by the Enterprises; the first column is only a reflection of the organizing principle behind the visualization. Anything you do, actively or passively (meta data) feeds many streams and purposes. Many more tentacles leave the middle column than feed it. Each piece of data collected feed many hungry mouths of the hydra.
In an ultimate irony, be forewarned that Google analytics drives the site. Which means that we cannot even create a site that looks at the data in the TOS without generating even more data.
Further details of their intent, methods and future plans can be found in the Mapping Data Flows Whitepaper. In addition, details of the team as well as the supporting institutions and foundations are also available in this paper.