Skip to main content
Epistemic News
LatestBlindspotArchiveSourcesPodcastSupport
Epistemic News

Every claim sourced. Every perspective labeled. News verified by AI, never edited by humans.

AboutMethodologyStatusPricingPrivacyTerms

© 2026 Epistemic News · Built by Wunderlich Web

Methodology

How we get closer to the truth — and where we fall short.

Source selection

We pull from 99 curated source feeds plus 23 Google News topic feeds, spanning wire services (AP, Reuters, AFP), US mainstream outlets across the political spectrum, international sources from 26 countries, and independent/nonprofit journalism. All feeds are polled every 30 minutes.

Source selection criteria: editorial standards, factual reporting track record, and spectrum coverage. We intentionally include outlets we disagree with — a truth engine that only reads sources it likes isn't seeking truth.

How bias ratings work

Each source is rated using Media Bias/Fact Check (MBFC) classifications. MBFC is the most comprehensive independent media bias database, rating 10,000+ outlets on both bias (far-left to far-right) and factual reporting (very high to very low).

We maintain inline bias ratings for all 99 sources in the feed list. Articles from Google News site-search feeds inherit the rating of their parent outlet. Sources not in our database default to "center" — this is a known limitation.

Important: Bias ratings describe the source, not the article. A left-center outlet can publish a perfectly neutral article. We label the source so you can judge framing — not to dismiss the content.

How clustering works

We generate vector embeddings for each article title using OpenAI's text-embedding-3-small model, then group articles with cosine similarity above 0.55 into clusters.

If embeddings fail, we fall back to TF-IDF (term frequency) matching on titles with a 0.35 similarity threshold. This is less accurate but ensures the pipeline never stops.

How synthesis works

For the top multi-source stories, we run a 4-agent adversarial debate:

Progressive analyst

Emphasizes humanitarian impact, institutional accountability, and systemic factors.

Conservative analyst

Emphasizes national security, fiscal responsibility, and traditional frameworks.

Libertarian analyst

Emphasizes individual liberty, government overreach, and market dynamics.

Devil's Advocate

Challenges all three — finds groupthink, missing angles, and logical gaps in every perspective.

A synthesis agent then produces the final article: consensus facts first, disputed claims clearly marked, blindspots noted, every factual assertion cited to its source.

How confidence scores work

Verified80+3+ independent sources agree, no contradictions.
Supported50–792 sources agree, minor framing differences.
DisputedBelow 50Sources actively contradict each other on this claim.
UnverifiedN/ASingle source only, no corroboration available.

Limitations and known biases

We believe transparency about limitations is more honest than pretending they don't exist.

  • △RSS-only ingestion means we miss stories that only appear on social media or behind paywalls.
  • △Some sources use Google News RSS proxies, which may return slightly different article sets than direct feeds.
  • △Sources not in our MBFC database default to "center" — unknown outlets get the benefit of the doubt.
  • △LLM synthesis can introduce subtle framing biases even with adversarial agents. The citations let you check.
  • △We synthesize 2-3 stories per run, not all 200+ clusters. Most stories pass through unsynthesized.
  • △English-language sources only. Global events are filtered through anglophone media.
  • △The pipeline runs on AI models (xAI Grok for synthesis, Anthropic Claude for ensemble voting, OpenAI for embeddings) that have their own training biases.