The U.S. is emerging from a hyper-polarized electoral season that boiled over in the storming of the U.S. Capitol by followers of President Trump on January 6. Ensuing efforts to hold the perpetrators accountable remind us that we live double lives: one in physical space, where our bodies act, and another online, where those actions are recorded – either by us or others.
Shortly after the mob broke into the Capitol, the FBI sought public assistance to identify the instigators. This type of crowdsourced public investigation is a novel phenomenon. The task is easier because some people uploaded videos and photos of themselves and others breaking into the Capitol to public social networks. Several of these people are already in custody.
Others incriminated themselves unwittingly. Leaked data from the Parler social network revealed over six hundred videos with GPS metadata that located users in and around the Capitol during the attack. A student quickly created a website that uses machine learning to sift through these videos to extract and publish individuals’ faces.
Such developments should concern all of us, regardless of where we are on the political spectrum or whether or not we abide by the law. Although Parler seems to have been remiss in its handling of geolocation metadata, it isn’t the only organization keeping track of its users’ whereabouts. A recent article on The Verge notes that the U.S. Defense Intelligence Agency (DIA) has bought such data from brokers without a warrant.
Another case of publicly available data helping identify the Capitol’s attackers — and one that is more directly relevant to those of us who structure information systems — involves the dating app Bumble.
Among other information, Bumble allows users to state their political affiliation. According to one Twitter user, at least one woman used this feature of the system to lure men who may have been involved in the attack, so she could report them to the FBI. The tweet drew a lot of attention, and Bumble eventually disabled its politics filter in the U.S. “to prevent misuse.”
I know a friend of a friend who changed her preference on Bumble to Conservative. She’s matching with MAGA bros and they’re bragging and sending her pics and videos of them in the Capitol. She’s sending them to the FBI.— Allison Norris (@allisonnorris) January 8, 2021
One can easily imagine why this field exists in a dating app. Few subjects are more contentious than politics, especially these days. People looking for a compatible mate would likely want to know the other persons’ political leanings. However, I doubt Bumble’s designers envisioned their system being used against its users in this way. (“The street finds its own uses for things.” — William Gibson)
I suspect most people don’t realize they’re not just beings with bodies that move in physical space, but also information ‘objects’ described (and identified) with metadata. The events of January 6 prove that the existence of these digital selves can have serious consequences.
While I’m pleased to see the Capitol’s attackers held to account for their actions, I’m also more aware of the implications of being a tagged individual in an information society. The same technologies and information structures that allow us to bring miscreants to justice can also be used for other, less noble, purposes.
Those of us who structure information environments must consider the implications of how we categorize and describe people. Information about people’s identity — including gender, race, nationality, religious commitments, and political affiliation — is especially sensitive.
Creating and maintaining categorization schemes bestows great power. People who design and manage such systems have tremendous ethical and moral responsibilities.
Icons by Maxim Kulikov from the Noun Project