Medium and GitHub scrub posts and database containing ICE employee LinkedIn data citing rules against doxxing; Twitter suspends bot tweeting info from database
Medium and GitHub say posting the LinkedIn data of a targeted group of employees constitutes doxing. The artist who made the database disagrees.
Context & Ripple Effects
The removals land on contested ground: LinkedIn has spent years fighting to control what happens to its members' data, including a lawsuit against 100 anonymous scrapers invoking the Computer Fraud and Abuse Act, yet individual profiles remain publicly viewable by design. Medium, GitHub, and Twitter are now drawing a line that turns on aggregation and targeting rather than source — the same profiles are fine alone, but a compiled database aimed at a group of ICE employees crosses into doxxing.
The dispute also sits inside a longer arc about ICE itself: internal documents detail hundreds of investigations into employees misusing its databases, which is precisely the accountability argument the artist's project was making. The platforms' call foreshadows how they would handle similar cases years later.
First-order effects
- The artist's database project is effectively shut down across three platforms at once — posts pulled from Medium and GitHub, and the bot tweeting from it suspended — leaving no major venue hosting the compiled data.
Second-order effects
- The doxxing rationale becomes a reusable template: X later suspends journalists and researchers over posting an alleged real name after Stonetoss appealed to Musk, and Meta blocks sharing of Ken Klippenstein's newsletter containing the JD Vance dossier, both extending the same move of removing targeted personal information at the subject's or policy's behest.
Third-order effects
- If aggregation-plus-targeting is the standard, the boundary of 'public data' is set by platform moderators case by case rather than by what any single profile exposes — a permission boundary that scrapers, researchers, and activists cannot verify in advance, while LinkedIn's own grip on its members' data remains separately contested.
The trend: Platforms are increasingly treating compiled public data about identifiable groups as doxxing, making moderator judgment — not technical accessibility — the real limit on what scraped information can be used for.