/
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

The New York Times begins using Alphabet's machine-learning powered Jigsaw to weed out toxic comments, increasing the number of articles that allow comments

Benjamin Mullin / Poynter :

Poynter Benjamin Mullin

Context & Ripple Effects

When Google's Jigsaw unveiled Perspective as a free API for publishers in February 2017, the pitch was simple: machine-scored toxicity would make comment sections affordable again. Four months later, The New York Times became one of the first marquee newsrooms to act on that pitch, adopting the tool and reopening comments on more of its articles.

The stakes run beyond one publisher: the same Perspective API later powered Jigsaw's Tune extension for filtering social-platform comments, was expanded in 2024 with finer-grained attributes, and now flags toxic output from LLMs built by OpenAI and Anthropic — making the NYT rollout an early proof point for a tool that became general-purpose content-integrity infrastructure.

First-order effects

  • The New York Times can open comments on more articles without proportionally growing its human moderation team, since Perspective pre-filters abuse before editors review queues.
  • Jigsaw gains a flagship publisher reference for Perspective at launch, strengthening its case to other newsrooms weighing whether to shut down or salvage their comment sections.

Second-order effects

  • Rival publishers face pressure to either adopt comparable ML moderation or accept that the Times can sustain community features they are cutting — turning moderation tooling into competitive table stakes rather than an experiment.
  • Reliance on automated scoring exposes publishers to misclassification risk: later testing found Perspective flagging innocuous phrases like "I am a gay black woman" as toxic, meaning early adopters inherit both the tool's efficiency and its blind spots.

Third-order effects

  • If the pattern holds, AI scoring becomes the default gatekeeper between readers and publishers, with human moderators supervising machine judgments instead of reading raw threads — a structural shift in who sets the boundaries of online discourse.
  • Perspective's trajectory from newspaper comment filters to a toxicity checker embedded in major LLM pipelines points toward a small set of shared moderation models underpinning both legacy media and frontier AI companies.

The trend: Newsroom engagement is shifting from abandoning comment sections to AI-assisted moderation, with Alphabet's Perspective evolving from a publisher utility into cross-industry toxicity infrastructure.