Skip to content
NewsTeam Documentation Request beta access

AI Features

NewsTeam’s approach to AI is simple: assist, don’t replace. We design AI to augment editorial judgment, speed up routine tasks, and enrich your content’s data model—always with human-in-the-loop controls and clear auditability.


  • Editorial-first — AI suggests; editors decide. You can accept, edit, or discard any suggestion.
  • Responsible by design — Clear handoffs, versioning, and activity logs preserve authorship and accountability.
  • Data-rich outcomes — AI helps build structured, reusable content (entities, locations, language, and more), powering search, personalization, and future products.

To find Content Wizard, Metadata Wizard and Analyze Article in Desk, see the article editor’s AI menu.

Analyze a draft and return a structured editorial review:

  • Score summary for title, synopsis, body, Originality & SEO, and total quality
  • Body feedback with corrections and concrete suggestions
  • Originality & SEO signals grounded in recent web research
  • Similarity risks where the angle appears close to common coverage
  • Source opportunities with facts, sources, or angles that would improve the piece
  • Additional insights with specific missing context to add

The analysis flow can run grounded web research, attach source links, and then keep a follow-up conversation open so editors can ask for rewrites, clarification, or deeper review of a specific section.

Generate or improve editorial building blocks:

  • Headlines & alternatives
  • Compelling intros
  • List-style summaries for quick takeaways
  • Social post options for distribution

Automate the time-consuming parts of enrichment—then fine-tune:

  • Keywords, tags, and social copy (synopsis, share text)
  • Language detection with ISO 639 codes
  • Mentioned Entities (people, companies, organizations, events)
  • Geo-tagging suggestions when locations are inferred from the story

Results are applied to the article’s structured metadata so content stays searchable, reusable, and consistent. See Articles → Metadata.

Create images with AI when appropriate. You can:

  • Generate an image from context (placement + surrounding content)
  • Edit or regenerate to match tone, placement, and usage
  • Keep full editorial control over what ships

Generate, refine, and update visual aids or infographic-style components from article context, instructions, or additional data. These outputs can include title, description, tags, mentioned entities, location context, and editable HTML.

Provide raw notes, a wire post, or a brief. AI will:

  • Produce a structured, publish-ready draft (with core metadata prefilled)
  • Suggest what to add next (gaps, missing voices, follow-ups)
  • Recommend widgets or visual elements to strengthen the piece
  • Add relevant non-text widgets, such as video embeds, when research surfaces useful source material

This is especially useful for turning early reporting into a solid first version.


  • Human-in-the-loop: Editors can review, accept, or modify every suggestion.
  • Provenance: All changes are captured via Revisions and the Messaging & Activity system for clear history and rationale.
  • Source visibility: Grounded analysis can return source links so editors can inspect the context behind recommendations.
  • Scoping: Bucket-level AI guidance can shape tone, locale, and editorial instructions for analysis and generation.