Practical Guide

Secure Communication for Journalists: A Practical Guide to Source Protection

May 8, 2026 9 min read Haven Team

A journalist covering sensitive stories faces adversaries that most people never encounter: intelligence agencies with mass surveillance capabilities, corporations with legal resources to subpoena communications records, and sometimes organized crime or foreign governments. The tools available to help are genuinely good — but they fail when the operational context around them is careless. Encryption solves one problem; knowing when to use which tool solves a different one.


The goal of source protection is that a person who provides information to a journalist cannot be identified from the journalist's communications, files, or devices. That goal encompasses at least three separate threat surfaces: the content of communications (what was said), the metadata of communications (who communicated with whom, when, and how often), and the physical or operational trail (who was seen where, who accessed what building, whose phone was near which cell tower).

End-to-end encryption addresses the first. It does relatively little for the second, and nothing for the third. Effective source protection requires thinking across all three.

The Threat Model Is Different from Ordinary Privacy

Most privacy advice is calibrated for a threat model that involves data brokers, advertising networks, and the occasional opportunistic account compromise. Journalists covering national security, organized crime, or corporate misconduct may face adversaries with substantially greater capabilities:

Tiered Communication for Different Sensitivity Levels

Not every source interaction requires the same level of precaution. Treating everything at the highest sensitivity level is operationally unsustainable and likely to lead to inconsistent practice. A tiered approach is more realistic:

Sensitivity Tier Examples Appropriate Channels
Tier 1 — Routine Confirmed officials, on-record sources, background interviews Standard encrypted email, Signal, phone
Tier 2 — Sensitive Confidential sources with some risk, document requests Signal with disappearing messages, encrypted email, news org VPN
Tier 3 — High-risk Whistleblowers in surveillance states, leakers in law enforcement or intelligence SecureDrop, air-gapped devices, in-person meetings, no digital trail

The tier determination should be based on the worst-case consequence if the source is identified — job loss, criminal prosecution, physical danger — rather than on your assessment of how likely exposure is.

SecureDrop for Anonymous Initial Contact

SecureDrop, developed and maintained by the Freedom of the Press Foundation, is the most widely adopted system for anonymous document submission to news organizations. The architecture provides strong protections for sources who need to make initial contact without revealing their identity:

SecureDrop has significant operational security requirements: sources must use Tor Browser (not a regular browser or VPN), ideally from a network not associated with them (not their home or work WiFi), and ideally from a device not associated with their daily use. These requirements are not incidental — they're what provides the protection.

Signal for Ongoing Source Communication

Signal is appropriate for ongoing communication with sources where you've already established contact and verified identity. Signal's end-to-end encryption means Signal Inc. cannot read message contents — but several characteristics affect how it should be used:

Enable disappearing messages. The default should be set based on sensitivity. For Tier 2 sources, one week is a reasonable default. For Tier 3, one day or less. Disappearing messages ensure that a device compromise or search some months later doesn't expose communications from before.

Verify safety numbers in person. Signal's safety number verification confirms that your communication is with the person you think it is and that no one is performing a man-in-the-middle attack. For sources whose identity matters, verify safety numbers through a second channel — ideally in person — before trusting sensitive conversations.

The phone number requirement is a real limitation. Signal requires a phone number for account registration. A source in a repressive environment may be identified from the mere fact that they have your phone number, or that a phone number in your contacts was in contact with a journalist. Consider using Signal with numbers obtained specifically for this purpose, registered to devices not linked to your main identity or the source's.

The ProtonMail Lesson for Journalists

In 2021, ProtonMail complied with a Swiss court order to log the IP address of a climate activist who had contacted journalists. Encryption protected the message contents; it did not protect the source's network identity. End-to-end encryption and legal compulsion of metadata are separate threat surfaces, and the marketing doesn't always distinguish them clearly.

Email When You Must Use It

Email is a hostile environment for source protection. Even with end-to-end encryption (PGP), the following are typically available to anyone with legal process served on the email provider:

For initial contact from a source who doesn't know your SecureDrop address, email is worse than SecureDrop in almost every dimension. For ongoing communication with a source already known to you, encrypted email — using a provider that stores minimal metadata and under a jurisdiction you trust — is workable at Tier 2, not Tier 3.

Haven provides encrypted email and chat under the same identity, which reduces the number of separate communication channels to manage — relevant for journalists who want their sources' encrypted email contacts and encrypted chat contacts to be the same address space rather than fragmented across multiple apps.

Device Security and Legal Compulsion

Physical device security has two relevant dimensions: what happens if the device is lost or stolen, and what happens if it's seized by authorities.

Full-disk encryption is non-negotiable. Every journalist's device should have full-disk encryption enabled with a strong passphrase. On modern iPhones and Android devices, full-disk encryption is on by default and tied to the lock screen PIN.

The PIN/passphrase versus biometric distinction matters significantly in legal contexts (see also our piece on biometric authentication risks). In most US jurisdictions, law enforcement can compel you to unlock a device with a fingerprint or face scan but cannot compel you to reveal a PIN or passphrase. For journalists who may be detained or whose devices may be seized, PIN-only unlock provides materially stronger protection.

Assume any device that crosses an international border may be examined. US Customs and Border Protection has broad authority to search devices at ports of entry without a warrant and without articulating suspicion. Some journalists travel internationally with clean devices and restore from a secure backup after crossing.

The Operational Security Gap

The most common source identification failures don't involve breaking encryption. They involve operational mistakes that the encryption never protected against:

Technical security is a necessary condition for source protection, not a sufficient one. The threat model must account for the full context: who knew the meeting was happening, what physical traces were created, what metadata is embedded in any documents received, and who at the news organization knows the source's identity or enough to narrow it.

The Freedom of the Press Foundation publishes detailed operational security guides for journalists and news organizations. These are worth reading in full for anyone working with Tier 3 sources. This post is a starting framework, not a complete operational security playbook.

Try Haven free for 15 days

Encrypted email and chat in one app. No credit card required.

Get Started →