Back to publications
Methodological demonstration OSINT methodology Version 1.0

Reference OV-PUB-2026-001

How to assess the reliability of information published online

A methodological demonstration of OSINT analysis based on cross-checking open sources

Published
July 31, 2026
Updated
July 31, 2026
Status
Public publication
Publisher
Intelligence Osiris Veritas Inc.

Executive summary

  • The repetition of a claim across several websites does not automatically constitute independent corroboration when those publications rely on the same original source.
  • An assessment should examine the origin of the information, the available evidence, the date, the context and confirmation from genuinely independent sources.
  • A rigorous analysis distinguishes directly verifiable observations, statements attributed to third parties, analytical inferences and elements that remain uncertain.

Question and scope

Analytical question

What criteria can be used to assess the reliability of publicly available online information in a structured way?

Period
General method applicable as of the publication date
Geographic scope
No specific geographic limitation
Content covered
Online publications, institutional websites, media and content distributed through digital platforms
Exclusions
Confidential data, closed systems, illegally obtained information and investigations targeting private individuals

Key methodological findings

Finding 1 — Repetition does not equal corroboration

Confidence: High
Observation
The same claim may be repeated by several websites, accounts or media outlets without adding new evidence.
Analysis
The number of repetitions should not be confused with the number of independent sources. The analyst must trace the claim back to its origin and determine whether the other publications rely on separate information.
Related sources
[1], [2]
Limitations
A secondary publication may sometimes add independent confirmation. Each repetition must therefore be assessed individually.

Finding 2 — The original source must also be assessed

Confidence: High
Observation
A document, statement or image issued directly by the relevant entity is generally closer to the event than a secondary comment.
Analysis
A primary source is not automatically accurate, complete or impartial. Its author, date, purpose, context and supporting evidence must be examined.
Related sources
[1], [2]
Limitations
In some cases, the original source may be inaccessible, deleted, modified or impossible to authenticate with certainty.

Finding 3 — A professional-looking website is not enough

Confidence: High
Observation
Graphic design, a domain name or a professional tone do not, on their own, establish the reliability of a webpage.
Analysis
Lateral reading means temporarily leaving the page to identify who is behind the source, what other reliable sources say about it and what independent evidence is available.
Related sources
[1], [2]
Limitations
Lateral reading depends on the availability of relevant and sufficiently independent external information.

Finding 4 — Authentic content can still be misleading out of context

Confidence: High
Observation
An image, quotation or statistic may be authentic while being associated with the wrong date, location or interpretation.
Analysis
Verification must address both the apparent authenticity of the content and the context in which it was created, published and reused.
Related sources
[1], [2]
Limitations
The absence of metadata or an original version may prevent a definitive conclusion.

Eight-step verification method

  1. Define the claim precisely and separate its different components.
  2. Identify the earliest publication or the most direct source that is accessible.
  3. Examine the author, organization, date, context, known funding and apparent purpose of the source.
  4. Determine what independent and recognized sources say about the author, website or claim.
  5. Assess the available evidence: original document, data, image, video, register, statement or direct observation.
  6. Compare dates, locations, versions, quotations and data to identify inconsistencies.
  7. Separate observed facts, reported statements, analysis and unconfirmed elements.
  8. Assign a confidence level and document the limitations before publication or transmission.

Evaluation grid

Criterion Question to ask
Origin Who first published the information, and can that version be located?
Identity Who is the author or organization behind the publication?
Proximity Did the source directly produce, observe or receive the information?
Date Does the content actually correspond to the claimed period?
Context Is the content presented with its full context?
Evidence What verifiable elements specifically support the claim?
Corroboration Are there genuinely independent confirmations?
Consistency Do the sources, dates, locations and data align?
Interest Could the source have a political, financial or reputational interest?
Limitations What cannot be confirmed, and why?

Application to a fictional scenario

Fictional claim

“A municipal service will be closed on Monday because of an emergency.”

Checks to perform

  • Locate the earliest known publication of the claim.
  • Determine whether it comes from an official municipal website or account.
  • Compare the publication date with the Monday mentioned.
  • Review official notices, calendars and public communications.
  • Look for independent confirmation without counting simple repetitions.
  • Keep open the possibility that the information is true, false, outdated or still unconfirmed.

Careful analytical wording

“The claim was identified in several publications, but they appear to rely on the same original source. No independent official confirmation was found at the time of verification. The status remains unconfirmed.”

General limitations

  • Reliability is not an absolute quality: a source may be accurate on one point and incomplete or biased on another.
  • The absence of public confirmation does not automatically prove that a claim is false.
  • Some information cannot be verified using open sources alone.
  • Pages, publications and metadata may be modified or deleted after consultation.
  • Confidence levels should be reassessed when new information becomes available.

Sources and references

  1. [1] UNESCO — Journalism, “Fake News” & Disinformation: Handbook for Journalism Education and Training

    Publication: 2018 · Accessed: July 31, 2026

    View source

    Role in the analysis: General framework on disinformation, fact-checking and verification of content distributed through digital media.

  2. [2] Digital Inquiry Group — Civic Online Reasoning — Civic Online Reasoning

    Publication: Date not indicated · Accessed: July 31, 2026

    View source

    Role in the analysis: Evaluation framework focused on source identity, available evidence and comparison with other sources.

  3. [3] Digital Inquiry Group — Civic Online Reasoning — Lateral Reading vs. Vertical Reading

    Publication: Date not indicated · Accessed: July 31, 2026

    View source

    Role in the analysis: Introduction to lateral reading as a method for evaluating online sources.

  4. [4] Verification Handbook — Verification Handbook

    Publication: Date not indicated · Accessed: July 31, 2026

    View source

    Role in the analysis: Practical reference on step-by-step verification of digital and user-generated content.

  5. [5] Library of Congress — Primary Sources and Personal Artifacts

    Publication: Date not indicated · Accessed: July 31, 2026

    View source

    Role in the analysis: Resource on examining, observing and contextualizing primary sources.

Validation and editorial responsibility

This methodological demonstration is published by Intelligence Osiris Veritas Inc. It presents general verification principles and does not constitute legal, financial or other professional advice tailored to a specific situation.

Any analysis applied to a real file should be scoped, supported by documented sources, accompanied by an assessment of limitations and human-validated before transmission or publication.