Documents, sources, and interviews with senior Facebook employees detail the process and the difficulty of moderating Facebook Live
used to stream the Christchurch attack—launched, the company realized it didn't have processes in place to handle the volume of bad material, like self-harm. Had to create tools for mods http://motherboard.vice.com/ ... http://twitter.com/... Matt Navarra / @mattnavarra : According to internal training docs for Facebook content moderators, they can- ‘Snooze’ stream, so it resurfaces to mods again every 5 mins to check if anything has developed- Ignore it + close report- Delete stream- Escalate stream to specialized team http://motherboard.vice.com/ ... See also Mediagazer
Context & Ripple Effects
The Motherboard reporting lands two days after the Christchurch attack was livestreamed on Facebook and spread across YouTube and Twitter despite takedown efforts. It reveals that when Facebook launched Live, it had no processes for handling high volumes of harmful material like self-harm streams, and had to build moderation tooling after the fact — a gap that persisted even after Zuckerberg's 2017 pledge to add 3,000 community operations reviewers following Live being used to broadcast murders.
The internal training docs, shared by Matt Navarra, show what moderators can actually do with a flagged stream: snooze it so it resurfaces every five minutes, ignore and close the report, delete it, or escalate to a specialized team. That menu is the company's answer to a real-time video problem its review pipeline was never designed for.
First-order effects
- Facebook's moderators are now working from an improvised playbook — snooze, ignore, delete, escalate — built reactively after Christchurch exposed that no process existed for volume attacks on Live.
Second-order effects
- YouTube and Twitter, already shown struggling to halt copies of the shooting video in related coverage, face pressure to match whatever real-time review tooling Facebook builds, since removal on one platform doesn't stop re-uploads elsewhere.
Third-order effects
- If live video keeps outpacing human review capacity, platforms will be pushed toward automated detection at upload time and possibly regulatory mandates for real-time moderation — shifting liability from post-hoc takedown to prevention.
The trend: Live streaming is forcing social platforms to move content moderation from reactive report-and-review toward real-time detection, with each broadcast atrocity accelerating the shift.