Anonymity for Protecting Sources
Source protection goes beyond hiding names: contact paths, reporting time, material content, backgrounds, metadata, and post-publication reactions can all identify sources.
When thinking about protecting sources, you may first imagine not naming them in the article.
But that is not enough.
Sources are inferred not only from names in the article text, but also from contact paths, interview dates and times, material content, photo backgrounds, file metadata, and reactions after publication.
Especially in small organizations, whistleblowing, local communities, politically sensitive topics, and stories related to workplaces or schools, candidates can narrow just from "who knew that information."
This article organizes which information to look at and in what order when protecting sources.
Sources Are Inferred From More Than the Article Text
The first thing to look at in source protection is not whether a name appears directly in the article text.
How many people know that information? When, where, and with whom did contact happen? Within what range were materials distributed? Who will be suspected after publication?
This perspective is necessary.
| Clue | Why the source is inferred |
|---|---|
| Information content | If few people know it, candidates narrow |
| Reporting date and time | Compared with work schedules, entry/exit records, and movement history |
| Place | People involved are inferred from the meeting place or region |
| Distribution scope of materials | Reveals who had the materials |
| Tone of quotation | Way of speaking and position remain |
| Photos and audio | Backgrounds, voices, reflections, and metadata become clues |
In anonymity, even if a name is removed, "the person who could release this information" remains.
That is why the article needs to be reread from the source's position before writing.
Choose contact paths according to source risk
In source protection, the choice of contact method matters.
If you contact a source through the workplace email they normally use, records remain with the organization. If you use a social media DM, logs remain with the platform. If you use a phone, call history and cell tower information are involved. Even if you meet in person, movement history, surveillance cameras, and payment records remain.
| Contact method | Information that remains | Caution |
|---|---|---|
| Workplace email | Send/receive history, attachments, search logs | Records are likely to remain with the source's organization |
| Personal email | Account, IP, device, email headers | Connects with real names or past accounts |
| Social media DM | Login state, account correlation, time | Records remain inside the platform |
| Encrypted messaging | Contacts, device, backups | Device and cloud settings become weak points |
| In person | Movement, entry/exit, payment, surveillance cameras | Real-world records remain even if nothing remains online |
A safer contact method changes depending on the situation.
For low-risk reporting, ordinary contact may be enough. For high-risk reporting, after consulting an editor in charge or a trusted specialist, it may be necessary to use a dedicated intake point or anonymous submission system.
SecureDrop is an open source system for media organizations and NGOs to receive anonymous submissions. It assumes access through Tor and metadata reduction, and is used by many media organizations for source protection.
URL : https://securedrop.org/
Tor Browser is used to make the source's IP address less directly visible to the destination.
URL : https://www.torproject.org/
However, even when using Tor or SecureDrop, the source may still be correlated through their device, files, behavior time, or submitted content. Tools are important, but the operation around them must also be considered.
Do not hide too little or too much about where materials came from
In reporting, you need to show the reliability of materials.
However, if you reveal too specifically where materials came from, the source is put at risk.
For example, writing "a document distributed at the department meeting on the morning of June 12, 2026" increases reliability. But if that meeting had few participants, the pool of possible sources narrows.
| Information disclosed | Benefit | Risk |
|---|---|---|
| Specific date and time | Readers can understand the facts more easily | The pool of attendees or people with access narrows |
| Department name | The location of the problem becomes clear | The source's affiliation is inferred |
| Document number | Evidentiary value increases | Can be compared against distribution records or management history |
| Original quotation | The meaning of the statement is conveyed accurately | Speaking style and internal terminology remain |
| Material image | Persuasiveness increases | Watermarks, margins, layout, and metadata remain |
Source protection does not mean weakening facts. It means choosing the necessary granularity.
Separate information needed to tell readers from information that only endangers the source.
Read from a third-party perspective before publication
The person who wrote the article becomes used to the clues.
Even if you think "this much will not reveal anything," people inside the organization, people in the region, or people involved may understand it.
Before publication, ask an editor who does not know the source or a trusted third party to read the article and check the following points.
- About how many people the pool of possible sources narrows to
- Whether dates, places, departments, and titles in the article text are too specific
- Whether the tone of quotations or internal terminology reveals a person
- Whether photos, PDFs, or audio contain metadata or background information
- Whether publication time or the way follow-up reporting is released makes the source look suspicious
This is not censorship.
It is a check for delivering necessary facts to readers while protecting sources.
Do not break protection through replies after publication
Source protection continues after publication.
After the article is published, if the reporter adds too much on social media, information hidden in the article text may appear. Clues about the source may also be released through replies to inquiries, added explanations, lectures, podcasts, and email handling.
| Post-publication action | Risk |
|---|---|
| Talking about background on social media | The timing or place hidden in the article text appears |
| Answering inquiries in detail | The reporting path or material scope leaks |
| Adding details in follow-up reporting | Information known only to the source increases |
| Getting heated during rebuttals | Unnecessary specific clues become easier to release |
| Replacing materials | Metadata remains in the new file |
Post-publication handling is also part of source protection.
Look not only at the article body, but also at surrounding statements by the same standard.
Look at real-world records too
In source protection, focusing only on online communication causes blind spots.
If you met in person, station movement, entry/exit, payments, surveillance cameras, and hotel or cafe usage history remain. If you spoke by phone, call history and carrier-side records also matter. Even if a name is completely withheld online, the source may be suspected when real-world behavior overlaps with the article publication timeline.
| Real-world record | Why to check it in source protection |
|---|---|
| Entry/exit records | Shows who was in a building and when |
| Payment history | Becomes a clue to meeting place and time |
| Surveillance cameras | Contact or movement may be recorded |
| Transit history | Unusual movement may remain |
| Call history | The person contacted and the time remain |
For high-risk reporting, include real-world behavior in the threat model as well as communication paths, materials, and article text.
Summary
Anonymity for protecting sources is not achieved only by withholding names.
Sources are inferred from information content, contact paths, interview dates and times, distribution scope of materials, tone of quotations, metadata in photos and files, and post-publication statements.
Choose contact methods according to the source's risk. Systems such as SecureDrop and Tor Browser can be useful in some situations, but tools alone do not complete source protection.
Before publication, check from a third-party perspective how far the pool of possible sources narrows. After publication, avoid releasing unnecessary clues through social media or inquiry handling.
Source protection is not work done at the end of an article by withholding a name.
It is an operation that continues from the beginning of reporting through after publication.
Related tools
ExifTool
An external resource related to this article. Open it only when it fits your situation and threat model.
Why it is listed: It can help with the article topic, but it is outside Anonymity Sense and should be checked before use.
URL : https://exiftool.org/
SecureDrop
An external resource related to this article. Open it only when it fits your situation and threat model.
Why it is listed: It can help with the article topic, but it is outside Anonymity Sense and should be checked before use.
URL : https://securedrop.org/
GlobaLeaks
An external resource related to this article. Open it only when it fits your situation and threat model.
Why it is listed: It can help with the article topic, but it is outside Anonymity Sense and should be checked before use.
URL : https://globaleaks.org/