News Made Clear · Loading…
The gold label identifies selected, extensively researched articles. It describes the research process; facts, analysis, opinion and uncertainty remain separately identified.
Produced by our full AI newsroom: research, evidence assessment, writing, editorial review and a separate factual integrity review.
Produced with in-depth AI research and editorial review, with sources and assessed claims available alongside the article.
An original opinion from the NMC newsroom, grounded in sourced facts and reasoned arguments. Editorial judgements, inferences and uncertainties are distinguished from verified facts.
News Made Clear aims to help people recognise misleading information and form their own judgement. We connect news with the sources, background and uncertainties needed to understand it.
Whether the source is a politician, company, government or influential social media voice, the speaker does not determine the conclusion. We examine the claim and the evidence. A statement may be supported, contradicted or insufficiently substantiated, even when that challenges a popular belief.
Our purpose is to correct false or misleading claims and supply missing context, not to campaign against a person or political position. A different opinion is not inherently false, and examining a statement does not mean judging the person sharing it.
An opinion article opens with a position and explains the facts and reasoning behind it. We call false claims false and expose misleading presentation, identifying what is omitted or distorted and why it matters. The strength of the judgement follows the evidence.
Facts remain distinguishable from our interpretation. An established error alone does not prove a deliberate lie or a hidden agenda. A conclusion that something is impossible must hold within the conditions stated. We include material contrary evidence and revise our judgement when new information warrants it.
Source links remain accessible. Two websites repeating one study are not independent confirmations. A timeline establishes sequence; it does not by itself establish causation.
AI can organise, summarise and analyse information. It can also make mistakes. Its use guarantees neither accuracy nor neutrality. We use AI to help prepare articles and analyses. Article publication details identify the preparation method; outstanding checks appear with the relevant claims and notes.
✓ means recorded AI assistance without a current human review. ✓✓ means an authorised editor explicitly reviewed this exact text and version. An additional discussion accent means comments exist; it does not mean the community verified the article as true. A revised version does not inherit approval automatically. Unknown provenance is labelled as such.
Investigation of fraud, corruption, crime or serious misconduct remains possible regardless of the actor involved. Publication is separate. Recognisable serious allegations are flagged for human review of evidence, counterevidence, unknowns, attribution and wording. This conservative triage does not detect every formulation. “X was accused by Y” differs from “X committed the act”. Insufficient evidence can leave a conclusion unresolved; strong evidence should not be artificially softened.
Readers can discuss articles and submit correction requests. Substantive objections deserve evidence-based responses. Popularity, likes and political disagreement do not determine truth. Reports are reviewed individually; reporting does not automatically remove a comment.
Publication history records updates, making changes and their reasons available for inspection.
A headline can invite curiosity and use mild irony. Its factual premise must hold. A question mark does not validate an unsupported accusation, and essential context belongs in the article itself, not only in later comments.
A firm conclusion may provoke disagreement. We do not deliberately amplify outrage to generate comments, and advertising interests, payments or traffic must not determine which facts we acknowledge. Commercial placements must be identifiable and separate from our assessment.
When challenged, we first identify the objection: a fact, a legal or moral judgement, missing context or our wording. We address the strongest substantive objection, including one that overturns our conclusion. Agreement with AI is not the measure of a successful discussion.
Some publication steps are automated. Systems can select and process sources, prepare and check drafts, publish articles that meet the configured editorial checks, and produce translations and updates. Publication does not imply that Hugo van Duijnen has personally read or approved each article before it appears.
Automated checks and comparison between AI systems are not independent human review. Article-level provenance and version history identify recorded review and substantive changes. AI can make factual, contextual and translation errors.
News Made Clear strives for factually accurate and impartial reporting. We make our editorial principles, use of AI and research methodology transparent.
On request, we cooperate with an independent audit of our working methods, within a scope agreed in advance and with protection for personal data and security-sensitive information. This may include examining our source selection, assessment of evidence, human direction and correction procedures.
Critical conclusions must follow from evidence. We aim to make it possible to assess whether we consistently apply the same standards of evidence, consider relevant counterevidence and correct established errors.
Audit enquiries: contact@newsmadeclear.com