Anonymous sources can help uncover vital information, but using them in journalism also involves risks. Here we look at when to use them and when to avoid them.
Every reader has seen the phrase: “a senior official, speaking on condition of anonymity” and asked why they can’t give their name? How are we supposed to believe someone who will only speak secretly?
It’s a fair question, and is the reason journalists must think carefully about using anonymous sources. They can be one of the most useful tools in reporting and one of the most easily abused.
Handled with discipline, they bring to light wrongdoing that would otherwise stay buried. Handled carelessly, they can ruin reputations, mislead the public, and damage a news organisation’s credibility for years.
And the sources often put themselves at significant personal risk – for which the journalist or news organisation shares responsibility.
Why anonymity exists
Many important stories begin with someone who knows something and is afraid to say it. The nurse who sees patients put at risk, the civil servant who watches a contract awarded to a friend of a minister, the employee who knows the safety reports were falsified: all of them have something to lose, but all may be motivated by the public interest to find a way to speak out.
The fears are real and varied. Some sources fear losing their jobs or being quietly sidelined for the rest of their careers. Some fear prosecution under official secrets or confidentiality laws. In some parts of the world, and on some issues like organised crime, they fear violence against themselves or their families.
Corruption, abuse and national security are the subjects where powerful interests have most to hide and the sources have most to fear.
If journalists could only report what people were willing to say under their own names, the public would hear mainly what the powerful wanted known. Famous investigations like The Pentagon Papers, Watergate and countless inquiries into institutional abuse depended on people who would speak only if their identities were protected.
The core principle: hidden from the public – not from the editor
As a first principle, anonymity should protect a source’s identity from the public but not from the editor.
A reporter who tells their editor, “I have a source, but I can’t tell you who,” is asking the news organisation to publish on faith. Serious outlets don’t do that, for three reasons.
- Accountability. If a story turns out to be wrong, someone must be able to explain how it went wrong. An editor who has never known the source cannot assess the risk, defend the story or learn from the failure.
- Verification. An editor who knows who the source is can ask the questions the reporter might not think to ask. Is this person really in a position to know? Does their account fit what we know about the organisation? Have they a reputation for reliability, or for settling scores? Do they have a conflict of interest which may affect what they are telling us?
- Legal liability. If a story is challenged, it is the publisher, not the individual reporter, that is sued for defamation. The organisation cannot defend a claim it cannot substantiate, and it cannot judge the strength of its defence if it doesn’t know who stands behind the claim.
Take two well-known examples. In 1980 The Washington Post published “Jimmy’s World,” a harrowing account of an eight-year-old heroin addict, and it won a Pulitzer Prize. The reporter, Janet Cooke, had invented the child. Editors had asked who he was and where he lived. Cooke refused to say, citing the need to protect a vulnerable source, and the paper, unwilling to push, went ahead. The prize was returned, and the paper’s reputation took a severe blow. A source who is invisible to the editor can be invisible for the worst possible reason: they may not exist.
And in 2018, The New York Times produced a podcast called “Caliphate” investigating ISIS. It followed the radicalisation of a young Canadian who said he became an executioner for ISIS. The podcast’s primary source was given a pseudonym, essentially making him an anonymous source. The podcast garnered publicity and awards, but the premise was a lie, and the central source later admitted his stories were a hoax. The paper’s editors reflected they had been too excited by what they appeared to have uncovered to undertake all the checks they should have made.
What the reporter must be able to tell the editor
When a reporter asks for permission to use an anonymous source, the editor should expect to hear three things.
- Who the source is. This means the real name, not a description or a nickname.
- What position or access they have. How does this person know what they claim to know? Were they in the room? Did they see the document, or hear about it from a colleague who did? A source who is merely repeating gossip, however senior, is not a source for a serious allegation.
- Why they are talking. Everyone has a motive. Some are noble: conscience, outrage, a sense of duty. Others are less so: a grudge against a boss, a rivalry, a desire to damage a competitor or to advance a political agenda. A motive doesn’t automatically disqualify a source, and some of the best sources are driven by anger. But the editor and reporter must understand it, because it affects how much weight the claim can bear and what must be checked independently.
Corroboration: one source is rarely enough
Most serious news organisations apply a simple rule of thumb. An anonymous claim, particularly a serious allegation about a named person or organisation, needs a second independent source or documentary evidence before it is published.
“Independent” is the operative word. Two people who heard the same story from the same third party count as one source, not two. Two officials from the same department who have talked to each other may be echoing a single account. Real corroboration comes from a different route to the same fact: a document, a data trail, a second person with first-hand knowledge, or physical evidence.
The stakes rise with the seriousness of the claim. A background briefing about how a government expects a vote to go requires less scaffolding than an allegation that a named individual has committed a crime.
The danger of relaxing this standard is clear from the run-up to the 2003 invasion of Iraq. Much reporting on Saddam Hussein’s supposed weapons programmes relied on anonymous officials and defectors whose claims were not independently verified. The New York Times later published an editors’ note acknowledging that some of its coverage had not been rigorous enough, and that information from sources it should have treated more sceptically had not been sufficiently challenged. Anonymity had given those claims an air of insider authority they did not deserve.
The BBC’s experience shows the same risk from the other direction. In May 2003 a BBC radio report alleged that the government had “sexed up” its dossier on Iraq’s weapons. The story rested heavily on a single anonymous source, the scientist and Ministry of Defence adviser Dr David Kelly. When his identity became known he was questioned by a parliamentary committee, and shortly afterwards he took his own life. The resulting Hutton Inquiry criticised the BBC’s editorial processes, and the corporation’s chairman and director-general both resigned. Subsequent public inquiries confirmed the broad thrust of the story – that the threat from Iraq had been exaggerated to justify invasion – but the original reporting had not been sufficiently robust and, relying on a single anonymous source, the BBC had been unable to back up their reporting.
The language of non-attribution
Journalists use a set of terms to describe the conditions under which information may be used. The trouble is that they are used inconsistently, and sources and reporters often mean different things by the same word.
- On the record: the source can be quoted and named.
- On background: the information can be used, but attributed only to a general description, such as “a senior government official” or “a person familiar with the discussions.”
- Off the record: in its strictest sense, the information cannot be published at all and is offered only to help the reporter understand. In practice, many people use it to mean “don’t name me,” which is a quite different thing.
- On deep background: the information can be used, but with no attribution to any source at all. The reporter must present it as their own knowledge, which is only defensible if it has been independently confirmed.
These definitions are not universal. A source who says “this is off the record” may think they have protected themselves completely, while the reporter hears “you can use it, just don’t name me.” That misunderstanding has ended careers and friendships.
The remedy is simple. Agree the terms explicitly, before the interview begins, in plain words rather than jargon. “I can use what you tell me, but I’ll describe you only as ‘a senior hospital manager.’ Is that acceptable?” is better than relying on shared shorthand understanding. And a source cannot retroactively declare something off the record after saying it; the agreement must come first.
The agreement between reporter and source
This conversation is the foundation of the whole relationship. Before the source says anything of substance, they should know exactly what protection they are being given: whether their name will be shared with an editor, how they will be described in print, whether and how documents will be used, and what will happen if lawyers or authorities come asking.
A good reporter explains this clearly and does not oversell. Sources who understand the limits of their protection can make an informed choice. Sources who discover them later feel, rightly, that they were misled.
Particular care must be taken if children are involved. Often it is left to parents and guardians to decide how identifiable or anonymous their children will be. However, this can lead to difficult questions about minimising future harm for children whose parents have given consent. For example:
- How will children feel seeing personal information published online?
- Will their family be publicly defined by what we write?
- Will the story appear in a Google search if a future employer looks up their child’s name?
Children can be particularly vulnerable and may have to live with the consequences later in life. Extra care must be taken to ensure anonymity.
Never promise more than you can deliver
The most important discipline in the agreement is restraint. A reporter should never promise absolute anonymity in circumstances where it cannot be guaranteed.
Legal systems differ on how far journalists may refuse to reveal sources. In the United Kingdom, section 10 of the Contempt of Court Act 1981 protects sources, but courts can order disclosure where it is judged necessary in the interests of justice or national security, or for the prevention of disorder or crime. In the United States, the Supreme Court held in Branzburg v. Hayes (1972) that the First Amendment does not give reporters a general right to refuse to testify before a grand jury, and protection varies from state to state.
Two cases show what is at stake. In 1990 Bill Goodwin, a young trainee reporter with The Engineer magazine, was ordered by a court to hand over his notes identifying the source of a leaked confidential document. He refused and was fined. The European Court of Human Rights ruled in 1996 that the order violated his right to freedom of expression, and the judgment became a landmark for source protection in Europe.
In the United States, the New York Times reporter Judith Miller spent 85 days in jail in 2005 for refusing to testify about a source in a leak investigation, and was released only when the source explicitly freed her from her promise.
Both reporters kept faith with their sources, but at considerable personal cost. A reporter who promises “I will go to prison before I name you” must mean it, and the source should understand that the legal risk is real.
Protecting sources in practice
Keeping a promise of anonymity takes more than good intentions. It takes practical care, much of it technical.
Communication. Ordinary email and phone calls leave records. Where there is real risk, reporters use end-to-end encrypted messaging, secure drop systems, or face-to-face meetings, and leave phones behind if location could be tracked. Press-freedom organisations such as the Committee to Protect Journalists publish detailed digital-safety guidance for this purpose.
The copy itself. Identifying details can reveal a source as surely as a name. “A senior finance director at the company’s Leeds office who joined in 2019” may describe exactly one person. Good reporters describe sources in terms broad enough to protect them but specific enough to tell readers why they are credible.
Documents and metadata. A leaked file can carry a hidden trail: author names, edit histories, printer tracking dots, or access logs showing who opened it. Images may contain location data. Reporters must examine, and where necessary strip or avoid publishing, anything that could lead back to the source. Showing a document in a story can expose who had access to it.
Storage and handling. Notes, recordings and contact details need to be kept securely, and the fewer people who hold the source’s identity the better.
When anonymity should be refused
Not every request for anonymity deserves a yes. Reporters should be willing to decline, and sometimes to walk away, when:
- The source wants to attack someone unfairly. Anonymity is for exposing wrongdoing, not for settling scores or making accusations that the accuser would not be willing to stand behind.
- The source wants to avoid accountability. If a claim could easily be made on the record by someone with nothing to fear, there is little justification for hiding behind a pseudonym. Powerful people routinely seek anonymity to say things they would not want to be held to.
- The source is trying to plant disinformation. Governments, companies and political campaigns all use anonymous briefings to float false claims or to test public reaction. A reporter who grants anonymity to a source with an agenda, without checking the claim, becomes a conduit for that agenda and is implicated in it.
- The information is available elsewhere. If the same facts can be established on the record or from documents, there is no need for anonymity.
A good test is to ask: what would the source lose if named, and does the public benefit justify the protection? If the honest answer is “very little” on the first and “not much” on the second, the answer should be no.
Editorial sign-off
Because the risks are so high, many serious newsrooms do not leave the decision to the reporter alone. The rule at many organisations is that a senior editor must explicitly approve the use of an anonymous source, and often must know the source’s identity, before the story runs.
This is not bureaucracy for its own sake. It spreads responsibility, ensures that a second experienced mind has examined the evidence, and protects reporters from the pressure of deadlines and their own enthusiasm. Where a story is especially sensitive, legal advice is often added to the process.
What the professional codes say
The principles above are not one newsroom’s preferences. They run through the leading ethical frameworks for journalism.
The Society of Professional Journalists’ Code of Ethics in the United States tells reporters to identify sources clearly so the audience can judge their reliability, to consider a source’s motives before promising anonymity, and to reserve anonymity for those who may face danger or retribution and who have information that cannot be obtained any other way. It also calls for explaining why anonymity was granted.
The Ethical Journalism Network, which promotes standards of accuracy, independence and accountability around the world, emphasises transparency about sourcing and the duty to give the public enough information to judge the credibility of what it is told.
Press-freedom organisations such as Reporters Without Borders (RSF) and the Committee to Protect Journalists (CPJ) approach the question from the other side, treating the confidentiality of sources as a cornerstone of press freedom and campaigning against laws and surveillance that put it at risk. They also provide practical safety guidance for reporters working in dangerous environments.
Taken together, these bodies are saying the same thing: protection of sources is a serious duty, and it must be conducted with rigour.
Why it matters
Readers are right to be wary of unnamed sources. Every anonymous quotation asks the audience to trust not just the person speaking but the journalist and the organisation behind them. That trust is fragile, and it can only be earned by showing that anonymity is rare, justified, checked and explained.
Done properly, it is among the most important things a free press can offer: a way for the powerless to tell the truth about the powerful without being damaged by it. Done badly, it is a licence to mislead and puts the source, the reporter and the news organisation at risk. The difference lies almost entirely in the discipline of the people involved.
