Bias Rating System

Every story on NewsReal.ai gets a bias tag and a manipulation score. Here’s what they mean and how they’re generated.

Bias Tags

Each story is tagged with a bias label reflecting the dominant framing of the source’s coverage — not the story’s topic, but how the outlet chose to present it. Tags include:

• Left-Leaning — framing favors progressive narratives or Democratic positions
• Right-Leaning — framing favors conservative narratives or Republican positions
• Establishment — framing supports institutional authority and status-quo positions
• Anti-Establishment — framing challenges institutional narratives regardless of left/right
• Corporate — framing protects or advances specific corporate interests
• Sensationalist — framing prioritizes emotional engagement over substance

Manipulation Score

The manipulation score (0–100) measures how aggressively a headline or story employs known persuasion techniques. Higher scores indicate more manipulation, not necessarily more bias. A story can be biased with a low manipulation score (straightforward advocacy) or highly manipulative with no clear political lean (pure clickbait).

Factors that increase the score:

• Emotional language designed to provoke rather than inform
• Missing context that changes the story’s meaning
• False equivalence or false dichotomy framing
• Buried lede — the most important information hidden or omitted
• Appeal to authority without substantiation
• Headline/body mismatch — when the article doesn’t support the headline

Important Caveats

These ratings are AI-generated assessments, not editorial judgments. The model evaluates framing techniques and language patterns — it does not determine truth or falsity. A story rated “low manipulation” can still be wrong, and a “high manipulation” story can still contain accurate information. Use these scores as one signal among many.