How to Safely Share Information Without Amplifying Disinformation
Sharing information can expose wrongdoing, preserve evidence, and help researchers connect important records. It can also unintentionally spread false claims, expose private people, reward manipulative accounts, or give misleading content a larger audience.
This guide explains how EpsteinWiki contributors can share verified information, correct falsehoods, and discuss uncertain claims without becoming a distribution channel for disinformation.
Snapshot
Before sharing anything, determine its verification status, public value, potential harm, and current reach.
| Information status | Recommended action |
|---|---|
| Verified and relevant | Share the verified fact with its primary source and necessary context |
| Important but unverified | Send it privately to the verification team or appropriate editor |
| False claim with little exposure | Avoid repeating it publicly unless correction is necessary |
| False claim with substantial exposure | Publish a concise, fact led correction supported by evidence |
| Manipulated or miscaptioned media | Preserve the original, label the problem, and display only what is necessary |
| Personal or sensitive information | Restrict access and apply EpsteinWiki privacy safeguards |
| Possible immediate safety threat | Escalate privately through the appropriate safety channel |
| Satire or parody | Label it clearly if there is a realistic risk of confusion |
A statement becoming popular does not make it reliable. Likewise, low engagement does not automatically make a claim harmless. The decision to share should depend on evidence, public interest, reach, and foreseeable harm.
Why Sharing Can Amplify False Information
Online platforms often treat interaction as a sign that content should be shown to more people. Replies, quote posts, reaction videos, reposts, tags, and outraged comments can therefore increase the visibility of the content being criticized.
A correction can also introduce a false claim to people who had never encountered it. Repeated exposure may make information feel more familiar, and familiarity can be mistaken for credibility. Research summarized by the Harvard Kennedy School Misinformation Review indicates that repetition requires careful handling because repeating both accurate and inaccurate statements can influence perceived accuracy.
Contributors should correct the record without turning EpsteinWiki into a distribution channel for the falsehood.
Potential amplification can occur through:
- Reposting or quoting the original content
- Embedding a misleading post in an article
- Repeating the false claim in a headline
- Displaying the claim prominently in a thumbnail
- Tagging the account that created it
- Linking directly to a manipulative account
- Publishing the creator’s username when identification adds no public value
- Replaying an entire manipulated video
- Using sensational wording that becomes searchable
- Repeating the same false statement throughout an article
- Adding misleading phrases to metadata or keywords
A correction should make the verified information more memorable than the falsehood.
Understand the Type of Information
Different information problems require different responses.
Misinformation
Misinformation is false or inaccurate information shared without necessarily intending to deceive.
A person may share an incorrectly dated photograph because they mistakenly believe it is current. The absence of malicious intent does not make the information accurate, but it may affect how a correction should be written.
Disinformation
Disinformation is false or misleading information deliberately created or distributed to deceive, manipulate, intimidate, or disrupt.
Intent is difficult to prove. Contributors should not label a person a disinformation operative without credible evidence supporting that conclusion. It is usually safer to describe the content and the observable distribution behavior.
Malinformation
Malinformation involves genuine information presented or distributed in a misleading, harmful, or privacy violating way.
Examples can include publishing an authentic address to encourage harassment, presenting an accurate quotation without its qualifying context, or exposing a survivor’s identity.
Unverified information
Unverified information has not yet been supported or disproved by sufficient evidence. Unverified does not mean false. It means the conclusion remains open.
Manipulated media
Manipulated media includes images, audio, video, documents, or screenshots altered in a way that changes their apparent meaning.
Miscaptioned media
Miscaptioned media may be authentic but falsely described. A real photograph from one date or location may be presented as evidence of an unrelated event.
Satire and parody
Satire may be misunderstood when removed from its original context. If a satirical item is relevant to an investigation, preserve the original context and label it clearly.
The American Psychological Association provides additional background on misinformation and disinformation terminology.
Verify Before You Amplify
The safest time to prevent amplification is before publication.
Find the earliest available source
Do not assume that the account displaying a claim created it. Search for:
- Earlier versions of the post
- The original document
- The complete video or recording
- The first known upload
- Official filings or government releases
- Earlier news coverage
- Archived versions of the page
- Corrections or retractions
Tracing a claim to its source may reveal that a quotation was shortened, a date was changed, or media was taken from an unrelated event. The Cybersecurity and Infrastructure Security Agency recommends tracing a message to its initial source before amplifying it.
Inspect the supporting evidence
Ask:
- Does the source provide evidence or merely repeat the allegation?
- Is the complete document available?
- Can the quotation be found in the cited source?
- Is the date correct?
- Does the record identify the person being discussed?
- Has the material been edited?
- Is the location independently verifiable?
- Are there other credible sources supporting the claim?
- Could the content be satire, impersonation, or engagement bait?
Use specific authorities
Avoid vague phrases such as “experts say,” “investigators confirmed,” or “documents prove” when the responsible authority can be identified.
Research published by the Harvard Kennedy School Misinformation Review found that misinformation may rely on nonspecific references to authority. EpsteinWiki articles should name the agency, court, document, researcher, or publication supporting the statement.
Pause before posting
Urgency is frequently used to discourage verification. Phrases such as “share before it is deleted” or “the media will not show you this” are not evidence.
The CISA Tactics of Disinformation resource recommends slowing down, checking facts, questioning sources, and thinking before linking.
Decide Whether Sharing Is Necessary
Not every false or questionable claim deserves a public response.
Before sharing, consider:
- Is the information relevant to EpsteinWiki’s mission?
- Has the claim reached a meaningful audience?
- Could silence cause identifiable harm?
- Would publication introduce the claim to a substantially larger audience?
- Can the correction be supported by primary evidence?
- Is the person being discussed a public figure, a private person, or a survivor?
- Could sharing expose identifying or sensitive information?
- Is the original content attempting to provoke engagement?
- Can the issue be addressed without linking to the originator?
- Would a private moderation action be more effective?
A false claim circulating among a handful of accounts may not require a public article. A false claim affecting survivors, court records, public safety, or the reliability of widely circulated evidence may require a visible correction.
Document the decision when the potential harm is significant.
Lead With the Verified Fact
Headlines, opening paragraphs, captions, and social previews should begin with what the evidence establishes.
Avoid using the false claim as the main headline, even if it is followed by a question mark or the word “false.” Readers may see only the preview, headline, or screenshot.
Weak approach
Secret document proves unidentified person was involved
Better approach
Reviewed court record does not identify the person named in circulating posts
The improved version states the verified conclusion first. It does not present the unsupported allegation as an open possibility.
A fact led correction generally follows this sequence:
- State the verified fact.
- Warn briefly that an inaccurate claim is circulating.
- Identify the claim only as specifically as necessary.
- Explain why it is inaccurate or unsupported.
- Restate the verified conclusion.
- Link directly to the supporting evidence.
Accuracy prompts can improve the quality of information people choose to share, according to research published by the Harvard Kennedy School Misinformation Review.
Describe the False Claim Only Once
A correction usually needs enough detail for readers to understand what is being corrected. It does not need to reproduce the falsehood repeatedly.
Use neutral language such as:
A circulating post incorrectly states that the document contains a particular name. The complete filing does not contain that name.
Avoid:
- Repeating the claim in every section
- Copying inflammatory language
- Embedding multiple copies of the same post
- Turning an unsupported allegation into a clickable headline
- Using the false statement in image captions
- Including unnecessary names or identifying details
- Speculating about the creator’s motives
- Mocking people who believed the claim
If the false claim must be quoted, quote only the minimum wording needed for identification. Clearly label it before readers encounter it.
Link to the Evidence, Not the Performance
Whenever possible, direct readers to the evidence that resolves the issue rather than the account promoting the controversy.
Preferred destinations include:
- Court records
- Government releases
- Complete source documents
- Full transcripts
- Original photographs or recordings
- Established archival captures
- Transparent correction notices
- Relevant EpsteinWiki evidence pages
- Epstein related records available through Epstein Data
For an EFTA record hosted by Epstein Data, use the complete document link in this format:
Replace the number signs with the record’s actual identifier.
Do not link directly to a disinformation account merely to show that the claim exists. If proof of the post is necessary, preserve an evidentiary copy and provide a carefully labeled excerpt. Direct linking may reward the account with traffic, engagement, followers, or search visibility.
Avoid Engagement Traps
Manipulative accounts may seek reactions rather than belief. They can benefit when critics argue with them, tag them, or quote their content.
High risk interactions
- Quote posting the false claim
- Reposting it with an angry caption
- Tagging the originating account
- Encouraging followers to confront someone
- Creating a reaction video that replays the entire item
- Replying repeatedly to provoke a public argument
- Using the same hashtags as a coordinated campaign
- Directing readers to search for the content
- Publishing screenshots that prominently display the account name
- Announcing that content is “being censored” without evidence
Safer alternatives
- Publish a separate fact led correction
- Link to the primary evidence
- Describe the issue without reproducing the entire item
- Blur or omit irrelevant account identifiers
- Archive the original privately
- Report platform violations through the proper process
- Send suspected coordination evidence to moderators
- Avoid instructing readers to engage with the source
Do not organize mass reporting campaigns or public confrontations. Reporting should be based on documented platform violations, not disagreement.
Evaluate the Claim’s Current Reach
Reach should affect the response.
Low reach claims
For a false claim with little exposure:
- Preserve it if it may become relevant
- Avoid public repetition
- Correct it privately when appropriate
- Monitor whether it spreads
- Do not publish the account merely to criticize it
- Record any connection to a coordinated campaign
High reach claims
For a claim already affecting a substantial audience:
- Issue a clear correction
- Lead with the verified fact
- Publish the strongest supporting source
- Avoid unnecessary links to the originator
- Correct captions, previews, and graphics
- Make the correction easy to share independently
- Monitor whether the false claim changes form
Visible likes, comments, and reposts are imperfect measures. A post may be viewed widely without receiving much engagement. Conversely, concentrated engagement may result from a relatively small group.
Describe observable metrics precisely. Do not claim that something “went viral” unless reliable evidence supports that description.
Write Safe Headlines and Social Previews
Many readers see only a headline, preview card, thumbnail, or search result. Each element must remain accurate when viewed without the rest of the article.
Headline rules
A headline should:
- State the verified conclusion
- Avoid sensational wording
- Distinguish facts from allegations
- Identify uncertainty
- Avoid repeating an unsupported accusation
- Use precise dates and document descriptions
- Avoid implying guilt through association
Preview and thumbnail rules
A preview should not:
- Display a false statement without a visible warning
- Use an unrelated photograph
- Place a private person’s face beside an allegation
- Present an edited image without labeling it
- Use clickbait language
- Suggest that a correction confirms the underlying rumor
Review automatically generated previews before publication. Social platforms may select an unsuitable image or truncate the qualification that makes a headline accurate.
Handle Screenshots and Cropped Documents Carefully
Screenshots can preserve useful evidence, but they can also conceal missing context.
Before sharing a screenshot:
- Find the original page or document.
- Record the complete URL.
- preserve the publication or posting date.
- Capture the surrounding context.
- Confirm that the screenshot has not been altered.
- Identify where the image first appeared.
- Archive the original source.
- Record the time of capture.
- Note any unavailable or deleted material.
- Remove unnecessary personal information from the public copy.
A cropped court filing should not be treated as a complete record. Locate the full document and check whether omitted pages, footnotes, docket text, or attachments change its meaning.
Do not use a screenshot of another person’s summary as a substitute for the underlying evidence.
Handle Manipulated Images, Audio, and Video
Manipulated media should be preserved for examination without being repeatedly distributed.
Preserve first
Keep:
- The original file as received
- The source URL
- The retrieval date and time
- Available metadata
- A cryptographic hash when appropriate
- The account and post identifiers
- The surrounding caption and comments
- Any earlier versions
- The verified original, if located
Verify independently
Check:
- Whether the media existed before the claimed event
- Reverse image and keyframe search results
- Lighting, shadows, reflections, and geometry
- Audio discontinuities
- Lip synchronization
- frame duplication
- Metadata inconsistencies
- Compression differences
- Whether an original recording is available
- Whether reliable witnesses or records corroborate the content
AI detection tools may assist triage, but a detector score does not prove that media is authentic or artificial. Record the tool, version, date, input, and limitations if a detector is used.
Display minimally
When correction requires showing manipulated media:
- Label it before display
- Use a still image when the full video is unnecessary
- Crop out private information
- Avoid autoplay
- Do not place it in a misleading thumbnail
- Provide the authentic source beside it when available
- Explain the specific manipulation
- Preserve the full item in restricted evidence storage
Never upload suspected child sexual abuse material or graphic sexual content to public or general contributor systems. Stop reviewing it and follow the appropriate reporting and safety procedure.
Handle Breaking News and Developing Claims
Information during a developing event may change quickly. Early reports often contain incomplete attribution, incorrect identification, or recycled media.
A developing article should separate:
What is confirmed
List facts supported by identified sources.
What remains unverified
Describe unresolved claims without presenting them as likely facts.
What has been disproved
State the verified correction and supporting evidence.
What changed
Provide a dated update explaining what was added, corrected, or removed.
Use precise timestamps when timing matters. Do not silently rewrite a material conclusion. Preserve revision history and add a correction note when an earlier version was inaccurate.
Share Information Safely in Groups
Local groups, volunteer channels, and private chats can spread misinformation rapidly. A message marked “private” can still be copied, forwarded, or captured.
Create a dedicated verification channel when possible. Information submitted there should include:
- The exact claim
- The original source
- The date found
- The contributor’s verification steps
- Supporting and conflicting evidence
- Any privacy concerns
- The proposed publication status
- The urgency level
Use clear status labels:
| Status | Meaning |
|---|---|
| Verified | Supported by sufficient reliable evidence |
| Partially verified | Some elements are supported, but important details remain unresolved |
| Unverified | Insufficient evidence to reach a conclusion |
| Disputed | Credible sources materially conflict |
| False | Reliable evidence directly contradicts the claim |
| Manipulated | Media or documentation has been materially altered |
| Miscaptioned | Authentic material is presented with false context |
| Restricted | Access is limited because of privacy, safety, or legal concerns |
Do not use labels such as “probably true” or “everyone knows” as substitutes for evidence.
Protect Survivors, Minors, and Private People
The public interest in correcting misinformation does not override basic privacy and safety duties.
Do not publicly share:
- A survivor’s identity when it is not already lawfully and responsibly public
- The identity of a minor
- Home addresses
- Private telephone numbers
- Personal email addresses
- School or workplace details that create a safety risk
- Medical or counseling records
- Graphic sexual descriptions that add no evidentiary value
- Images that permit visual identification of protected people
- Unverified accusations against private individuals
Redact identifying details before using screenshots or documents. Check filenames, metadata, thumbnails, alt text, captions, and linked attachments because private information may remain in those locations after the visible text has been removed.
Do not speculate about a person’s identity from blurred images, body shape, clothing, voice, or resemblance.
Use Prebunking When Appropriate
Prebunking prepares readers to recognize common manipulation techniques before a specific false claim spreads.
A prebunking notice may explain that readers should watch for:
- Cropped documents without docket information
- Old images presented as new
- Fake quotations overlaid on photographs
- Impersonation accounts
- Urgent requests to repost without verification
- Anonymous claims presented as insider knowledge
- Screenshots with no original link
- Artificially generated audio
- Coordinated accounts repeating identical wording
- Real documents paired with unsupported conclusions
Focus on the technique rather than predicting or repeating a specific allegation. The purpose is to build recognition skills without giving an emerging falsehood a ready made script.
The World Health Organization discusses prebunking and other approaches intended to strengthen resistance to misinformation.
Correct Mistakes Promptly and Transparently
EpsteinWiki contributors must correct their own errors with the same care used to correct external misinformation.
A correction should:
- State what was wrong.
- Provide the accurate information.
- Identify the supporting evidence.
- Explain when the correction was made.
- Update the headline, body, caption, and preview.
- Preserve the revision history.
- Notify affected editors or channels.
- Remove unsafe copies when appropriate.
- Avoid silently replacing a material allegation.
- Review related articles for the same error.
Publish the correction through the channel where the original error reached readers. A correction hidden on a separate page may not reach the affected audience.
The Society of Professional Journalists Code of Ethics emphasizes verification, context, original sources, prompt corrections, and minimizing harm.
Suggested Correction Language
Verified information
Verified fact: The complete court record states [accurate summary]. The relevant document is available here: [primary source].
Unverified claim
We are aware of this claim. It has not been verified and should not be repeated as fact. EpsteinWiki is reviewing the original source and available supporting records.
False claim
Correction: [verified fact]. A circulating post incorrectly states [minimal description of the claim]. The complete record shows [evidence based explanation].
Miscaptioned image
This photograph is authentic, but the circulating caption is inaccurate. The image was originally published on [date] in connection with [verified context].
Manipulated media
This media has been altered. The verified original is available at [source]. The alteration changes [specific feature or meaning].
Removed material
We removed an earlier version because the available evidence did not support its conclusion. The revision record has been preserved.
Sensitive information
This material contains identifying or sensitive information and will not be reposted publicly. A redacted evidentiary copy has been preserved for authorized review.
Developing information
This is a developing matter. The following facts are confirmed as of [date and time]. Other circulating claims remain unverified.
Safe Sharing Workflow
Use this workflow before publishing or forwarding potentially controversial information.
Step 1: Preserve the original
Save the URL, content, date, time, context, and available metadata.
Step 2: Identify the claim
Write the claim in neutral language. Separate observable facts from conclusions.
Step 3: Locate the source
Find the original document, post, recording, photograph, or filing.
Step 4: Check authenticity
Examine provenance, completeness, dates, media integrity, and source identity.
Step 5: Seek corroboration
Look for independent evidence that does not merely repeat the same origin.
Step 6: Classify the information
Mark it verified, partially verified, unverified, disputed, false, manipulated, miscaptioned, or restricted.
Step 7: Assess public value
Determine whether sharing advances an investigative, educational, archival, or safety purpose.
Step 8: Assess harm
Consider privacy, harassment, defamation, survivor safety, minor protection, and the risk of further amplification.
Step 9: Evaluate reach
Determine whether the claim is contained, emerging, or already widespread.
Step 10: Choose the least amplifying format
Use a separate correction, primary evidence link, limited excerpt, or private escalation as appropriate.
Step 11: Review every surface
Check the headline, Snapshot, body text, captions, graphics, tables, links, related guides, metadata, and search description.
Step 12: Publish with a status and date
State the verification status and the last meaningful review date.
Step 13: Monitor the result
Watch for new evidence, altered versions of the claim, reader confusion, and unintended exposure.
Step 14: Correct the record
Update promptly when credible evidence changes the conclusion.
What Not to Do
Do not:
- Share first and verify later
- Repeat an allegation because it is trending
- Treat engagement as proof
- Use a screenshot as the sole source when the original is available
- Name a private person merely because an account speculated about them
- Present AI detector output as conclusive
- Contact or confront suspected participants
- Encourage readers to harass, report, or investigate someone
- Publish personal information to disprove a rumor
- Download or redistribute unlawful sexual material
- Use sensational headlines for corrections
- Hide uncertainty in footnotes
- Delete material errors without a correction notice
- Describe speculation as an investigative lead without labeling it
- Assume that a blue check, official looking logo, or large following proves authenticity
- Link to manipulative content when a primary source can establish the relevant fact
- Declare coordinated activity without documenting the observable pattern
- Infer criminal involvement from contact, proximity, a photograph, or inclusion in a record
Final Sharing Checklist
Before publication, confirm:
- The original source has been located or its absence is disclosed.
- The information has an accurate status label.
- The headline begins with the verified fact.
- The false claim is described only as much as necessary.
- Primary evidence is linked.
- Dates and locations are accurate.
- Images, recordings, and screenshots have been checked.
- Automated detection results are not presented as proof.
- The content does not expose survivors or minors.
- Private information has been removed from text and files.
- The originator is not being rewarded with unnecessary traffic.
- The article does not encourage confrontation or harassment.
- Uncertainty and conflicting evidence are visible.
- Social previews remain accurate when viewed alone.
- Corrections and revision history are preserved.
- The sharing decision serves a clear public interest.
- The article remains understandable without sensational language.
- The SEO description does not repeat the false claim unnecessarily.
If several of these conditions cannot be satisfied, do not publish yet.
Questions
Should I repost a false claim so people know it is false?
Usually not. Publish a separate, fact led correction and link to the evidence. Repost the original only when readers must see a limited portion to understand the correction.
Does adding “false” to a repost prevent amplification?
No. The interaction may still increase visibility, and readers may encounter the original claim without reading your correction.
Should a low reach rumor receive a public response?
Only when the risk of harm or likely spread justifies introducing it to a wider audience. Quiet correction, documentation, or monitoring may be safer.
What if the claim is already viral?
Publish a clear correction promptly. Lead with verified information, explain the evidence, and avoid directing additional attention to the originating account.
Can I use a screenshot instead of linking to the original post?
Sometimes, particularly when a direct link would reward a manipulative account. Preserve the original URL and context, label the screenshot, and remove irrelevant personal information.
Should usernames be visible in correction screenshots?
Only when identification has clear evidentiary or public interest value. Blur or omit usernames belonging to private or low reach accounts when naming them would create unnecessary exposure.
Is an anonymous tip safe to share if it sounds credible?
No. An anonymous tip should be verified independently. Credibility is not established by detail, confidence, urgency, or claimed insider access.
Can several posts count as independent confirmation?
Not if they trace back to the same source. Repetition by multiple accounts is not independent corroboration.
What if a real document is being used to support a false conclusion?
Link to the complete document, explain what it does and does not establish, and identify any omitted context.
Should EpsteinWiki speculate about why someone created a false claim?
No, unless credible evidence establishes motive. Describe observable content and behavior instead.
Can I call an account a bot?
Only when sufficient technical and behavioral evidence supports the classification. Repetitive activity alone does not prove automation.
Can AI detection tools authenticate an image or recording?
No. They may provide investigative signals, but their outputs can be inaccurate and should not be treated as conclusive proof.
What if I already shared something inaccurate?
Stop further distribution, preserve the original version, publish a prompt correction, update every affected surface, and notify relevant editors.
Should the incorrect article simply be deleted?
Deletion may be appropriate for unsafe material, but a material error should normally have a visible correction record. Preserve the editorial history in the appropriate system.
Can a private group message still amplify disinformation?
Yes. Private messages can be forwarded, copied, screenshotted, or repeated publicly. Apply the same verification discipline.
How should rapidly changing information be labeled?
Use a visible developing status, state what is confirmed, identify what remains unresolved, and add dated updates.
What if the correction could expose a survivor?
Do not publish identifying information. Correct the broader factual issue without naming or indirectly identifying the survivor.
Is it acceptable to link to an archived copy?
Yes, when the archive is reliable and the link is necessary for verification. Warn readers if the archived page contains false, graphic, or sensitive material.
Should the false claim appear in SEO keywords?
Generally no. Including it may increase its search visibility. Use factual subject terms unless identifying the false claim is essential to a high reach correction.
When should moderators be contacted?
Contact moderators when content presents a safety risk, exposes protected information, appears coordinated, violates platform rules, or requires restricted review.
What if credible sources disagree?
Label the issue disputed, present each source accurately, explain the conflict, and avoid forcing a conclusion the evidence cannot support.
Related EpsteinWiki Guides
- Disinfo and Manipulated Media Guide
- Fact Checking and AI Detection Tools
- How to Recognize Online Manipulation, Bots, and Infiltrators in Local Groups
- How to Vet Tips, Sources, and Whistleblowers
- How to Archive Web Pages and Social Media Posts So They Hold Up as Evidence
- EpsteinWiki Investigative Standards
- Epstein Files Evidence Framework
- Handling Contradictory Evidence
- Contributor Safety Guide
- Privacy Safeguards for Minors
- Volunteer Conduct Code
- Style Guide
Sources
- CISA: Tactics of Disinformation
- CISA: Disinformation Stops With You Toolkit
- CISA: Think Before You Link
- CISA: Disinformation Stops With You Infographic Set
- Harvard Kennedy School Misinformation Review: Expert Views on Misinformation
- Harvard Kennedy School Misinformation Review: Accuracy Prompt Toolkit
- Harvard Kennedy School Misinformation Review: Evaluating Misinformation Interventions
- World Health Organization: Combatting Misinformation Online
- World Health Organization: Immunizing the Public Against Misinformation
- American Psychological Association: Misinformation and Disinformation
- Society of Professional Journalists Code of Ethics
- UNESCO: Journalism, Fake News and Disinformation