AI News Claim Checker Workflow: Webz.io + GPT-4.1 to Sheets

Turn news questions into checkable, source-backed claims—automatically

This AI News Claim Checker workflow (Webz.io + GPT-4.1-mini) searches recent news, converts results into distinct factual claims, labels each claim as corroborated or single-source based on outlet counts, and archives everything to Google Sheets for audit-ready review.

What this workflow does

• Receives a question from an n8n chat trigger.
• Classifies the request as a quick lookup or deep research task to control how many Webz.io searches run.
• Searches news with Webz.io and uses GPT-4.1-mini to extract and group coverage into distinct claims with supporting source URLs.
• Deduplicates sources by registrable domain and counts independent outlets per claim.
• Applies a corroboration threshold to label each claim as corroborated or single-source.
• If claims are found, it generates a sourced briefing (linked outlets plus a source-spread footer) and appends one row per claim to Google Sheets.
• If no usable coverage is found, it responds with suggested next search angles instead of producing an unsourced answer.

Use cases

• Verify whether a trending headline is supported by multiple independent news outlets.
• Build an internal “claims tracker” for analysts, SaaS operators, or comms teams.
• Turn ad-hoc research questions into structured claim/status/source logs in Google Sheets.
• Use quick vs deep research routing to manage Webz.io search costs and latency.

Technical details

• Integrations: Webz.io (Bearer Auth + news search tool), OpenAI GPT-4.1-mini, Google Sheets
• Workflow logic: if, set, code, merge, split out, sticky note
• Sheet columns: saved_at, question, claim, status, source_count, publishers, urls
• Includes configurable research settings such as default language, lookback days, and corroboration threshold.