/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

Twitter says it removed 50% of abusive tweets in Q3 before users even flagged them, up from 38% in Q1, credits its moderation algorithms

Igor Bonifacic / Engadget :

Engadget Igor Bonifacic

Context & Ripple Effects

This caps a two-year enforcement ramp: back in mid-2017 Twitter said it was taking action against 10x the number of accounts it did a year earlier, and by that September it was suspending most terror-related accounts before their first tweet ever posted. The Q3 2019 figure extends that pattern from account-level discipline to tweet-level interception.

What changed is where the work happens: instead of acting after a report or an account accumulates strikes, the company now says its moderation algorithms catch half of abusive tweets unprompted — the same proactive-detection logic Facebook later cited when it attributed its bullying-and-harassment takedown growth to reviewer capacity and AI improvements.

First-order effects

  • Users see fewer abusive tweets before they are flagged, shifting the burden of discovery from victims' reports to Twitter's own systems — the metric itself (50% vs 38% in one quarter) becomes the evidence the enforcement is working.

Second-order effects

  • Rivals face a new benchmark for transparency reporting: once Twitter publishes a proactive-detection rate per category, peers like Facebook are pushed to disclose comparable pre-report figures, turning moderation metrics into a competitive scoreboard.

Third-order effects

  • If platforms compete on pre-flag removal percentages, moderation investment becomes a standing cost of operating a social network rather than a reactive fix — and regulators gain a ready-made quantitative yardstick to hold platforms against.

The trend: Content moderation is moving from reactive, report-driven takedowns toward algorithmic pre-publication interception, with platforms publishing detection rates as proof.