Are anonymity-seekers just like everybody else? An analysis of contributions to Wikipedia from Tor

TL;DR

The paper analyzes Tor-origin edits to English Wikipedia to quantify their value when anonymity-seeking users are blocked. Using 11,363 Tor edits from 2007–2018 and time-matched controls (IP, First-time, Registered editors), the authors examine reversion rates, persistent content, manual labels, and ORES classifications to assess quality. They find that Tor edits are generally similar in quality to non-Tor unregistered edits but exhibit higher revert and controversy activity, with topic preferences distinct from other editors. After 2013, improvements in blocking reduce Tor edits, yet the study suggests that allowing controlled anonymity-enabled contributions could benefit Wikipedia and society, prompting design options beyond outright bans. Overall, the work provides a data-driven perspective on the trade-offs of blocking anonymity and offers evidence to inform policy and tooling improvements for trusted anonymous participation.

Abstract

User-generated content sites routinely block contributions from users of privacy-enhancing proxies like Tor because of a perception that proxies are a source of vandalism, spam, and abuse. Although these blocks might be effective, collateral damage in the form of unrealized valuable contributions from anonymity seekers is invisible. One of the largest and most important user-generated content sites, Wikipedia, has attempted to block contributions from Tor users since as early as 2005. We demonstrate that these blocks have been imperfect and that thousands of attempts to edit on Wikipedia through Tor have been successful. We draw upon several data sources and analytical techniques to measure and describe the history of Tor editing on Wikipedia over time and to compare contributions from Tor users to those from other groups of Wikipedia users. Our analysis suggests that although Tor users who slip through Wikipedia's ban contribute content that is more likely to be reverted and to revert others, their contributions are otherwise similar in quality to those from other unregistered participants and to the initial contributions of registered users.

Figures (8)

• Figure 1: Screenshot of the page a user is shown when they attempt to edit the Wikipedia article on "Privacy" while using Tor.
• Figure 2: Number of edits per month by Tor users to English Wikipedia between 2007 and 2018.
• Figure 3: Reversion rate for edits from different groups of editors over time (2008--2013).
• Figure 4: Measurement of PTRs of different groups of non-revert edits over time. The rectangle is the interquartile region (middle 50% of the population), with an orange line at the median. The upper and lower whisker represent the range of the population that is 1.5 times above and below the interquartile range. The green triangle is the mean, and the circles indicate individual observations falling outside the limit.
• Figure 5: A non-parametric LOESS curve over time. We use feature injection to instruct the ORES Good Faith model to treat all edits as if they were made by a newly created user account.