After several weeks of renewed scrutiny over its “real name” policy, Facebook recently announced that that it intends to “fix how this policy gets handled.” As advocates for the benefits of online anonymity and the use of pseudonyms, we’re encouraged to see Facebook in a dialogue about their real-name rule and glad to see them be responsive to people who have been negatively impacted by the policy. The ability to define one’s own identity is closely intertwined with the rights to freedom of expression and access to information. The Internet supports an unprecedented capacity for anonymous and pseudonymous communication, enabling people to explore ideas, information, and opinions without the chilling scrutiny of their local communities or the world at large. People use pseudonyms to do everything from ask potentially embarrassing questions to challenge their governments.
People also create their own identities that are just as “real” as anything linked to a government document. This was a point of contention in the latest dispute over Facebook’s account name policy, as a number of drag queens, trans people, and others who live – and use social media – under names of their own choosing found their accounts suspended for violating Facebook’s real-name policy. As Sister Roma, of the drag group Sisters of Perpetual Indulgence, noted after being compelled to change her Facebook account name, “[my legal name] is not how I identify myself or how anyone identifies me.”
As Facebook works to update its account name policy to better respond to the needs and interests of its diverse user-base, CDT has some thoughts on key issues it should consider.
When social media companies talk about “real names”, it’s never completely clear what they mean. Individuals like Chinese journalist and political blogger Michael Anti have faced account deactivation on Facebook for not using the name on their government-issued identity cards, even when they have extensive records of writing and publishing under their chosen name. This has led many advocates to frame Facebook’s policy as a “legal name” requirement, to highlight the fact that Facebook seemed to be relying on what government documents declared a person’s name to be, as opposed to what that individual considered their actual name and identity. Facebook pushes back on this term in its announcement, stating that the “policy has never been to require everyone on Facebook to use their legal name.” This is a welcome clarification, and should guarantee that Sister Roma and others affected in this latest identity sweep get their accounts reinstated. But we’re eager to see Facebook further specify what their policy actually is.
The announcement notes that, “The spirit of our policy is that everyone on Facebook uses the authentic name they use in real life.” This calls to mind the former policy of Google Plus, which instructed users to “use the name your friends, family or co-workers usually call you.” (In a welcome move earlier this year, Google dropped that vague policy and no longer places any restrictions on usernames.) The drawback of these attempts to describe a “real-ish name” policy is that they rely on terms like “real life” and don’t account for the fact that one person can present themselves in multiple ways in the course of living their real life – just think of the differences in how you interact with your boss, your best friend, and your mother. Online, it can be all too easy for spheres of our lives to overlap when we’d prefer they remain separate. Pseudonyms are one crucial tool that people use to shore up the separation of these spheres on social media.
So Facebook will need to further clarify just how much pseudonymity they’re prepared to support. Facebook should take this opportunity to consider the ways in which its real-name policy affects people in a variety of circumstances all over the world. Different Facebook users will have a variety of reasons for wanting to maintain accounts under different sorts of names. Is this “authentic name” policy limited to full-time identities or nicknames? Does it allow someone to maintain parallel professional and personal accounts? Will it permit activists who use online platforms to engage in advocacy with deeply significant real-world influence to see Facebook as a viable option for their work? Facebook has a real opportunity now to make meaningful changes to its name policy and enforcement procedures, and to ensure that these polices are clearly communicated to its community of users.
How to authenticate someone’s identity once an account has been flagged is another challenge that many companies face. Documents containing significant amounts of personal information, such as a driver’s license, are highly sensitive, and companies should think carefully before taking on the responsibility to be a steward of that data. As Facebook develops its authentication procedures, it must take utmost care to protect the security of any information it obtains from users and delete it as soon as authentication has been completed. And it should go without saying that these documents should not be linked to or incorporated into Facebook’s profiles of its users.
As for authentication when a person’s account name is not the same as the name on their driver’s license, Facebook should approach the issue thoughtfully and be very clear in its explanations once it arrives at a new policy. Facebook notes that authentication documents might include a “gym membership, library card, or piece of mail.” To be sure, these documents would be much less sensitive than a driver’s license, and could enable some users to authenticate in a way that makes them feel safe and their identity respected.
But these alternative forms of ID may be just as difficult to obtain under a person’s preferred identity; for example, to obtain a library card in the DC library system, a person needs to “present two current forms of traceable identification, one to prove identity and one to prove place of residency.” Clarity about what alternate forms like “a piece of mail” is essential, as well – does this mean a letter from a friend, a magazine, or a utility bill? Many less-official types of identifying documents are issued only after a person provides an official credential; if the alternate methods of authentication that Facebook will accept all ultimately trace back to a driver’s license or apartment lease, they may not really provide much of an alternative. And Facebook has users from across the globe, so an ideal set of authentication standards would be sensitive to the different (and often complicated) contexts surrounding identity and authentication that exist in different cultures.
But community flagging systems are also vulnerable to abuse; as Facebook notes in its announcement, the people who were singled out in this most recent spate of real-name enforcement were identified because “[a]n individual on Facebook decided to report several hundred of these accounts as fake.” That is, someone used Facebook’s name policy and flagging system to target and try to silence people they suspected were using something other than their legal names.
Facebook’s announcement is a clear statement that the company doesn’t want to facilitate this sort of silencing, which is an important first step. Now, Facebook will need to conduct a thorough review of its flagging system and identify the potential vulnerabilities it contains. In our Account Deactivation and Content Takedown paper with the Berkman Center, we identified a number of safeguards that platforms should incorporate into any content moderation system, including:
- Require users who flag inappropriate content to specify the rule the content allegedly violates – don’t let users make vague complaints that leave content moderators guessing.
- Provide for intermediate steps or an escalation process during the content or account review process – don’t jump right from receiving a flag to deactivation.
- Accompany warnings, content removal and blocking, and account deactivations with immediate notice to the affected user, clear explanation of the violation, and descriptions of next steps – don’t leave people wondering why their account has been flagged.
- Develop an appeals process to mitigate the impact of mistakes or abuse by third parties – don’t leave users without recourse if someone else tries to silence them.
Facebook’s content moderation process already incorporates some of these recommendations, and we’re glad Facebook has shown a willingness to respond to the real-life needs of some of its users who want to use Facebook under names of their own choosing. The initial clarification that the real-name policy will embrace some of these users is a positive step, and we hope Facebook will further broaden its embrace to include the many different kinds of people who are clamoring for pseudonymity. Figuring out how to implement an account naming policy that supports pseudonyms while still allowing for robust anti-spam and other abuse prevention procedures can be a challenge – but one that’s well worth it for users’ freedom of expression and participation in online life. We hope Facebook will keep the conversation going, provide more clarity in their policy, and consider broader reforms to protect some of the most vulnerable people that use the Internet – and those who simply want to express who they really are.