BiasChecker.ai

The Questions Behind Every Analysis

Last updated: 26 October 2026

Every analysis is the same article read under a fixed set of instructions. This page is the plain-English version of those instructions: the expert role we give the AI model and the questions it sets out to answer for each analysis type. It’s a more detailed companion to our methodology overview.

Ground rules every analysis shares

The same handful of rules sits behind all the analyses below, so we state them once here rather than repeat them under each one:

  • Just the article text. We read the article's written text — the menus, ads and cookie banners around it are ignored. Images, video and audio aren't extracted either, so anything carried only by them falls outside the analysis.
  • It describes the writing — it can't judge people. The analysis states plainly what the writing does, but an automated reading of text has no way to establish a real person's guilt, motives, or inner state — so it never claims them. Findings describe the article and the conduct it reports, stay conditional about anything unverified, and are never accusations against the journalist or anyone named.
  • Verbatim, verifiable evidence. Findings must quote the article exactly (short fragments), and we check every quote against the source text so you can judge it yourself.
  • No scores from the model. The model never gives a numeric score or level; our code computes every verdict from what was found.
  • It trusts the article on recent events. The model's training has a cut-off date, so when the article describes something more recent, the analysis goes with the article rather than "correcting" it from possibly outdated knowledge.

↑ Back to top

What each analysis asks

One per analysis type. Each gives the model a different expert role and a different set of questions to answer about the same article.

Intent Analysis Intent

What the text says vs what it wants you to conclude — with a crisp summary of the piece

We ask the model to read the article like an editor and separate what the text literally says from what it steers you to believe or do, treating intent as a property of the writing.

The questions it asks

  • What does the text literally claim or report — its surface message?
  • What conclusion, belief, or action is the reader actually being guided toward?
  • Which techniques — word choice, emphasis, structure, whose voice gets heard — do that steering?
  • Does the headline match what the body actually supports — or is the gap between a punchy headline and a more careful body itself part of the steer? In the extreme, is the headline a lure for a body about something else entirely?
  • Is there anything buried in the text that quietly undercuts the steered conclusion?
  • What is one simple thing you could check in the text yourself to test the read?

↑ Back to top

Persuasion

8 persuasion techniques, named only on a clear fit

We ask the model to examine how the writing persuades you — including where it tips into manipulation — by surfacing the hidden assumptions it asks you to accept and the conclusion it leads you toward.

The questions it asks

  • Does the article’s own writing steer the reader beyond what its facts support at all? Straight reporting usually does not, and then there is nothing to name.
  • What is the reader expected to take as true without proof?
  • What conclusion is the piece leading the reader toward?
  • Which moves get the reader there — and does each one clearly fit a named technique, checked against the article’s own text?

Each technique below has its own explainer, with what the analysis must find in the article before naming it and two examples: persuasion techniques explained.

The techniques below are a vocabulary, not a checklist: each comes with a check the model must fill from the article’s own text, and a move that only loosely fits is left out. The article’s overall lean — loaded language, framing direction, selective emphasis — is deliberately not in this list: that is the Bias analysis’s job:

Placement and normalising — how the text arranges and softens

  • Buried fact — A fact that changes how a claim reads is placed where readers are likely to miss it.
  • Normalising — Something serious or disputed is presented as ordinary, popular or settled without support.

Reasoning — how the text gets you to a conclusion

  • Guilt by association — Something is judged by what the text places it beside, with no reason given for the link.
  • Unrepresentative example — A broad conclusion rests on examples or figures that do not represent the whole picture.
  • False choice — Only two options are presented when more exist.
  • Deflection — Instead of answering a criticism, the text turns to another subject, often the critic's own faults.
  • Assumed conclusion — A disputed conclusion is treated as already true instead of being argued for.
  • Emotional appeal — The text's own wording uses fear, pity, outrage or hope in place of argument.

A tactic is only flagged when the article’s own voice performs it — rhetoric inside a quote the article reports belongs to the speaker, not to the article, and quoted emotional testimony from named people is reporting, never the article’s own appeal to emotion. Open slant is not covert persuasion: a comedy piece, a warm profile, or plainly promotional copy (a press release, an organisation’s own announcement) wears its slant on its sleeve, so its favourable selection of facts is expected, and the usual correct output for such pieces is an empty list.

↑ Back to top

Bias Analysis Bias

29 bias categories

We ask the model to measure how much, and in which direction, the piece leans — pointing to the specific words and structural choices behind the tilt, across our catalogue of 29 bias categories.

The questions it asks

  • Which way does the piece lean, and which perspective does it favour?
  • Where do word choice, emphasis, or structure tilt the framing?
  • Is the sourcing one-sided, or are obvious perspectives left out?
  • When casualty or other figures for more than one side are reported, is the depth of coverage proportional?

The analysis scans for all 29 categories below — grouped by how much impact each tends to have — and reports only the ones the article’s own narration genuinely shows:

High impact

  • Framing — How information is presented to steer interpretation (includes implicit and agenda-setting bias).
  • Political — Political leaning or partisanship (includes nationalistic framing).
  • Racial — Race-related bias in coverage or language.
  • Gender — Gender-related bias in coverage or language.
  • Religious — Bias for or against a religion; faith-based assumptions or anti-religious framing.
  • Commercial — Undisclosed sponsorships, affiliate links, or financial interests.

Moderate impact

  • Source — Over-reliance on a narrow set of sources (includes authority bias).
  • Confirmation — Reinforcing an existing belief or narrative.
  • Selection — Cherry-picking data, quotes, or examples — including a slant-creating gap, where a missing perspective tilts the story (a full what’s-missing audit is the separate Omissions analysis).
  • Emotional — Loaded language and appeals to emotion.
  • Sensationalism — Exaggeration, clickbait, hyperbole.
  • Stereotyping — Generalisations about groups (includes socioeconomic).
  • Headline — Misleading or sensational headlines and titles.
  • False balance — Giving equal weight to unequal positions (bothsidesism).
  • Citation — Selective citation of supporting studies while ignoring contradicting ones.
  • Survivorship — Focusing on successes while ignoring failures.
  • Recency — Overweighting recent events over historical context.
  • Anchoring — Over-reliance on the first information presented.
  • Cultural — Ethnocentric or Western-centric perspectives.
  • Statistical — Misleading use of statistics; reading correlation as causation.
  • Age — Ageism and generational stereotypes.
  • Negativity — Systematic overemphasis on threats, failures, and bad outcomes.
  • Geographic — Urban-vs-rural or regional dismissal and stereotyping.
  • Disability — Ableism and problematic portrayals of disabled people.
  • Narrative — Forcing events into a hero/villain or rise/fall arc that distorts reality.
  • Access — Only covering stories journalists can easily reach.
  • Automation — Over-trusting AI or algorithmic outputs as fact.
  • Bandwagon — “Everyone believes X” appeals to popularity.
  • Just-world — Implying victims deserved their fate.

We assess only the article’s own narration — not inflammatory quotes it reports — and treat factual context, skeptical framing, and “contacted for comment” as signs of balance, not bias. The analysis first identifies what kind of piece it is: visible comic tone and openly promotional copy (a press release, an organisation’s announcement on its own site) are open slant, not hidden leans — a promotional piece that reports its facts accurately has its lean named at most once, never restated under several categories. The same rule against double counting applies generally: one lean gets one primary category, not the same tilt filed under framing, narrative, emotional, and sensationalism at once.

↑ Back to top

Omissions

What a well-informed reader would expect but the article leaves out

We ask the model to act as an investigative editor and run a dedicated “what’s missing” audit — the relevant, consequential, knowable context the article leaves out.

The questions it asks

  • What fact, context, perspective, counterpoint, or caveat would a well-informed reader expect to see?
  • How does each absence change the impression the article leaves?
  • Is each gap distorting — its absence changes the conclusion a reader draws — or merely useful background a fuller piece would add?
  • Does the headline fairly reflect the article — or does it overstate, contradict, or assert a certainty the body itself doesn’t support?
  • Is one party singled out for a practice that others in the story also take part in, by a route the article never mentions?
  • Where the article says a party can switch, replace, or find an alternative — does it state the condition that decides whether it really can?
  • Where would a reader look to fill the gap — what kind of source, rather than a specific link?

Gaps are judged against the article’s own topic and claims, so a thorough, balanced article can come back with few or none. Anything the article says, even inside a quote, counts as present, and a second model checks each claimed gap against the article: a gap is removed when that model quotes the sentence that covers it. Events after publication, and forecasts or knock-on effects that have not happened, never count as missing, and the article’s account of where a process stands is taken as given (though the lens may still ask whether the next step will really happen). Where the article clearly attributes a one-sided account (“prosecutors allege…”), the other side is usually treated as background. A headline is flagged only when it claims something the body does not support, not when it merely simplifies.

↑ Back to top

Critique

Tests whether the article’s argument, or a plan it reports, holds up: what it argues, where it fails, and a better approach. Straight reporting simply passes.

We ask the model to act as a domain expert and judge whether the article’s argument — or an open plan, policy, or agreement it reports — holds up, and if it does not, what would work better.

The questions it asks

  • What does the article argue, and on what main reason and evidence?
  • Is the goal the right one, and is the approach a sound way to reach it?
  • Which flaws would bring down the main claim, a point it rests on or the proposal, and which are minor?
  • Where a fair reader could repair the argument, what version of it would hold up?
  • What better course could the people involved realistically take, or, where the case holds, what other course would work as well?

It judges substance, not style: whether the reasoning holds up if the article’s facts are true, read in its strongest fair form. Persuasion tricks are left to the Persuasion lens, and a missing detail counts only when it would change the conclusion. Most articles are judged on their own argument. When an article reports a plan, policy or agreement that is still open, the lens judges the case made for that decision instead. Straight reporting is not judged at all: events, natural processes, acts of violence or force, crimes, lawsuits and court cases not yet decided, and completed history. A flaw is key if it would bring down the article’s main claim, a point that claim rests on, or what the article proposes; any lesser weakness is minor. The verdict depends on how many key flaws there are. When there is a key flaw, the lens suggests a better approach; when there is none, it may still offer another approach that works as well with different trade-offs. Every suggestion is something people could actually do.

↑ Back to top

Moral Lens Moral

Assess against moral principles

We ask the model to act as an applied-ethics analyst and judge whether the conduct is wrong — even where it is perfectly legal — looking separately at how the text presents events and at the events themselves.

The questions it asks

  • Does the text itself cross an ethical line — dehumanising language, mocking suffering, dishonesty?
  • Are the actions or events described morally problematic — proportionality, unnecessary harm, rights, exploitation?
  • Are the official justifications actually morally sufficient, or merely asserted?
  • Which specific moral principles are genuinely at stake here?

It weighs the article against a fixed set of moral principles in two groups — how the writing treats its subjects and readers, and whether the conduct it describes is justified — reporting only the ones genuinely at stake:

In the writing — how the article presents events

  • Dignity — Degrading or dehumanising language in the article’s own voice.
  • Truthfulness — Dishonesty or deception.
  • Compassion — Mocking suffering or celebrating tragedy.
  • Equality — Portraying groups as unequal in worth.
  • Wellbeing — Needlessly nihilistic; crushing hope without cause.

In the conduct described — the events themselves

  • Proportionality — A response out of all proportion; collective punishment.
  • Harm — Actions causing unnecessary harm without justification.
  • Justice — Unfair consequences or lawlessness.
  • Sanctity of life — Treating life as disposable.
  • Rights — Infringement of human rights.
  • Autonomy — Coercion, or restricting freedom without cause.
  • Responsibility — Avoiding accountability for the actions taken.
  • Privacy — Invasion of personal boundaries.
  • Property — Theft or destruction of possessions.
  • Fairness — Unfair treatment or distribution.
  • Exploitation — Taking advantage of vulnerable people.

A finding must tell you something the article doesn’t already: natural events with no moral agent are not condemned, conduct that appears only in a passing background clause is scene-setting rather than this article’s events, and conduct the article itself presents as settled wrongdoing is treated as the story, not a discovery — though where the text contests, defends, or minimises the conduct, the lens’s own judgement does add information and is given. The headline sentence — for example “the reporting is fair; the events it describes raise ethical concerns” — is computed from the two axes, so neutral coverage of grave events is never presented as a flawed article.

↑ Back to top

Scientific Assessment Science

Check scientific accuracy and rigor

We ask the model to act as an expert in scientific method and assess the rigor of both direct scientific claims and any reporting on research studies.

The questions it asks

  • Are the claims factually correct against settled scientific consensus?
  • Is the evidence appropriate and strong enough for the claims being made?
  • Are cause-and-effect claims actually supported, or is correlation being read as causation?
  • For science journalism: does it represent the study’s findings, limitations, and methods accurately — or overstate them?
  • What evidence would actually settle a genuinely disputed claim?

Recent findings are not flagged as errors merely because they fall past the model’s knowledge cutoff, and clearly-labelled religious, moral, or philosophical claims are not judged as failed science. The scope test cuts both ways: an article making no scientific claims at all is marked not-applicable rather than “scientifically sound”, while a press release or op-ed asserting an empirical or causal claim in its own voice is fully in scope — genre never exempts the claim.

↑ Back to top

Background

Topic context: what this is, how it developed, and where it stands now

We ask the model to act as a reference editor and write a neutral briefing for someone who has landed on the story mid-stream — the factual history of the topic itself.

The questions it asks

  • What is this topic, and where does it stand today?
  • How did it develop — the key milestones, oldest to newest?
  • Who are the main parties or roles involved?
  • Which parts are settled history, and which are recent or genuinely contested?
  • Where the public record allows: what was originally forecast, promised, or projected about this topic — and how did that actually turn out?

It stays neutral — describing how each side characterises a dispute rather than taking a side — and flags low confidence rather than inventing dates, quotes, or statistics. Relative time (“recently”, “on Friday”) is resolved against the article’s own dateline, not today’s date. The timeline stays strictly anchored to the article; the one exception is an optional “what was promised versus what happened” note, where the model may add well-documented forecasts or pledges from the public record that the article itself omits — included only when it genuinely recalls concrete specifics (a named institution, a pledged figure, a dated forecast).

↑ Back to top

Historical Parallels Parallels

Find historical precedents

We ask the model to act as a historian and find one to three past events that work the same way as the situation in the article: analogies, not a timeline of the topic itself.

The questions it asks

  • What historical event or period resembles this situation, and how closely?
  • What patterns match — and what is importantly different?
  • How did the historical case actually play out, and what does that suggest here?
  • What lesson does it offer?

It looks first in the history of the article's own field, such as science for a science story. Every article gets its nearest parallel: when that is a loose match, the analysis says how loose rather than overstating it. It may close with a fitting quotation, with a named source and any doubt about its wording or authorship stated.

↑ Back to top

Neutral Rewrite Rewrite

The condensed read: the article’s meaning separated from the spin, rewritten neutrally

We ask the model to act as a plain-spoken sub-editor and rewrite the article neutrally and more briefly: keep the material facts and each side’s position, strip out the loaded language, emotional padding, and persuasion — including the persuasive structure (ordering, emphasis, framing) — and add nothing new.

The questions it asks

  • Which words are doing persuasion rather than reporting — and what is the neutral way to say the same thing?
  • What is emotional staging, repetition, or filler that can be cut without losing a material fact?
  • Is the article’s architecture itself steering the reader — a countervailing fact buried at the end, one side given far more room, or the narration framing events as threats or reassurances — and how should the rewrite reorder and rebalance to undo that?
  • Is every claim still attributed exactly as the original attributed it, rather than promoted into a flat statement of fact?

This lens only removes: it never adds a fact, a figure, context, or a fact-check — anything not in the original article is left out. It may reorder and rebalance (leading with the most consequential facts, compressing an over-weighted side harder, moving a buried caveat next to the claim it qualifies), and when it strips a slanted sentence it must keep the facts that sentence carried — but it cannot restore perspectives or sources the original left out. It is also a rewrite, not a summary: a lean, factual piece should come back near full length, and phrases such as “The article notes that…” are removed — the output is the article, not a description of it. Direct quotations are used sparingly and only word-for-word; speech is otherwise reported indirectly. The model also names the story’s main actors, each verified verbatim against the article before display. An automatic check then compares the rewrite against the original: a quotation or accusation that is not in the source blocks the rewrite outright, and a number that cannot be verified against the original is shown highlighted rather than silently trusted.

↑ Back to top

Roast

A warm, witty comedic take on the article

We ask the model to act as a warm, sharp stand-up comedian doing a bit ABOUT the article — its one job is to be genuinely funny about it, not to grade the journalism.

The questions it asks

  • What punchy one-line tagline captures the article — the comedic hook?
  • How would you playfully retell what the article actually says?
  • Which specific details — a phrase, a claim, an absurd number stated with a straight face — are most worth a standalone joke?
  • What is the one honest, humour-free takeaway that grounds the bit?

The humour only ever targets the writing and framing — never the people in the story or what they are going through. Real tragedies (deaths, serious harm, disasters) and sensitive human hardships (illness, disability, grief, addiction) are declined outright, and a dry, cleanly-written piece with no comedic angle gets nothing to roast.

↑ Back to top

Synthesis

An analyst’s report on the article, written from its other analyses: what it’s doing to its reader, how, what checks out, and how to read it

The model works only from the article’s other analyses and adds no analysis of its own.

The questions it asks

  • In one sentence, what is the article doing to its reader: leaning, steering, or just informing?
  • Which few connected points explain how the piece works on its reader, and which findings back each one?
  • What does the article do well, and which analyses found nothing to flag?
  • What one thing should a reader check before making up their mind?

How it chooses its inputs

  • Built on the four core analyses. Synthesis always uses Intent Analysis, Persuasion, Bias Analysis and Omissions. The setup card runs any the article does not have yet, and you can add other analyses too.
  • Whose results it uses. For each input, the result from the model you chose for the synthesis; if there is none, a Consensus result; otherwise the most recent result from any other model. The report records which model each input came from.

Every point must cite findings from the inputs; our code drops any it cannot trace back, so quotes always come from the inputs themselves. The report is dry and to the point. It comments on the writing and its effect, and any passage our defamation check rejects is left out. It is not a fact-check: it reports what the analyses found and does not check the article’s claims against other sources. If the analyses found little, you get a short summary with no report.

↑ Back to top

↑ Back to top

Consensus: merging a panel

Consensus

Several models from different makers analyse the article; a separate model merges what they agree on.

Each panel model first analyses the article under that analysis type’s own instructions, described above. The consensus editor then writes the merged result. It never reads the article, cites only quotes the panel already cited, and adds nothing of its own.

The questions it asks

  • Which findings did at least two models report independently, even in different words? Our code counts this before the editor starts.
  • Where two models gave the same passage different labels, is it one issue or two? One issue is reported once, under the label that fits the evidence best.
  • What is one clear wording for each agreed finding, rather than a stitch of the models’ own phrasings?

How the panel works

  • Who is on the panel. One model from each of at least three makers (Anthropic, OpenAI, Google, DeepSeek, Z.ai, Moonshot AI, xAI). By default these are Claude Haiku 5.5, GPT-6 Luna and DeepSeek V3.2, and you can swap a model or add a maker before the run. Models that have already analysed the article join for free, one per maker, so a well-analysed article can have four or five at no extra cost. Gemini 3 Flash merges the results. We review this line-up as models change.
  • What you see. Each agreed finding shows how many models backed it, for example "3 of 4 models". Findings from a single model are listed under Raised by one model, with the model's name and evidence; usually the others were looking at other parts of the article, not disagreeing. If the models agree on nothing, or split on the overall verdict, the result says so.
  • Scored the same way. Our code verifies the quotes and computes the verdict exactly as for a single model, and the result records which models took part. If fewer than three models answer, the run reports the failure so you can try again.
  • Strengthen. A saved consensus does not change on its own. If the article later gains analyses that would make a stronger panel, a Strengthen button shows which models it would add or upgrade and what that costs, usually just the merge.

↑ Back to top

A living method

These instructions are refined continuously — sharper guardrails, clearer questions, new checks — so the wording shifts over time. The commitments above (no model-emitted scores, verbatim evidence, and never an accusation about a person) are stable and survive every revision.

Once an analysis runs, the result leads with a verdict — a quick read of what it found. The methodology overview explains the verdicts and what their levels mean.

↑ Back to top