What We’re Building Next: A Browser Extension for Instant News Analysis + a White-Label API
A few weeks ago I shared the story behind Neutral News AI, how it started with frustration at biased, fragmented news and turned into a project focused on transparent, evidence-based journalism.
Today I want to go deeper into what the product actually does and what’s coming next.
1. The Core Idea
Not “neutral news.”
Not ideology.
Just transparent, multi-source evidence.
Our system takes any news topic, then:
Collects coverage from multiple reputable outlets (left / center / right).
Extracts all atomic claims from the combined articles.
Retrieves evidence for each claim.
Runs MNLI fact-checking to classify each as: Supported / Contradicted / Inconclusive.
Generates a concise, source-bounded summary with all citations visible.
Everything is traceable.
Every sentence links back to its evidence.
If something is uncertain, we show it.
This is the part that early testers said “finally looks like journalism built on receipts.”
2. What we’re building now (the fun stuff)
A) Browser Extension (coming soon)
Paste nothing. Click nothing.
You read an article → click the NNAI icon → instantly see:
Political Bias & Bias Strength
Credibility & Trustworthiness
Sentiment & Emotions
Subjectivity
Readability
Clickbait Detector
And a summary for every article
Basically an “X-ray mode” for any news article.
This is the feature that most testers have asked for, and we’re already prototyping the UI.
B) B2B / White-Label API
Several journalists, analysts, and creators asked for programmatic access to our pipeline.
So we’re developing a clean API that lets anyone:
Submit a URL → get structured claims back
Retrieve multi-source evidence packs
Run bias detection, sentiment, credibility signals
Get MNLI fact-check results
Pull neutral summaries with citations
Retrieve comparison data across outlets
We want to support media outlets, researchers, fact-checkers, and even YouTube/podcast creators who want to provide more transparency in what they publish.
3. Why this matters
Most “fact-checking” tools stop at scores or flags.
Most “AI summary” tools hallucinate or remove context.
We built something different:
A transparent pipeline that shows exactly why a conclusion was reached.
No black boxes.
No ideology.
Just structured evidence.
4. Ask for feedback
I’d love to hear from the community:
Would you use a browser extension like this?
What output format would you want from an API?
Any specific features you’d like in a multi-source news analyzer?
Your feedback helps shape the roadmap more than you think.
Thanks for reading, and thanks to everyone who commented on the first post.
