Do Press Embargoes Still Hold?

What an embargo actually is

The cleanest published definition comes from the organisation that runs the largest embargoed-material pipeline in the world. EurekAlert!, the science news distribution service operated by the American Association for the Advancement of Science, defines a news embargo as “a prohibition against making public, in any way, certain information before a specific date/time”.

The words “in any way” are doing real work. An embargo is not only a bar on publishing an article. It covers posting the finding on social media, briefing it to a colleague at another outlet, and putting a headline behind a paywall that a scraper will find. In the version of the practice that actually functions, the embargo covers the information rather than the article.

The mechanism underneath is a trade. A source hands over material before it is public — a study, a results statement, a product, a report — and in exchange the recipient agrees not to use it until a stated moment. The source gets coordinated, simultaneous coverage instead of a scramble. The journalist gets time to read the thing properly, call an independent expert and write something that is not a rewritten press release. The theory, which is a good theory, is that this produces more accurate journalism than a free-for-all would.

Note what is not in the definition: any statement of what happens if the agreement is broken. EurekAlert! describes the obligation in detail and the penalty not at all, saying only that it “recognizes or enforces embargo guidelines set by legitimate content owners” — which places enforcement with the journal or institution, not with the pipe. That gap runs through the whole subject and is the reason this guide is built from cases rather than rules.

Who honours one, and why

Embargoes hold best in science and medical journalism, and the reason is structural rather than cultural. In that field, access is gated and repeated. A health reporter needs the next press package, the next conference press room, the next set of author interviews. A source that can withdraw all of that has leverage that a company issuing one product announcement does not.

EurekAlert! shows the gate explicitly. Reporters must digitally sign an embargo agreement before they can access the reporter-only section of the site, and public information officers must sign one before they can obtain an account to post embargoed releases. Both sides of the exchange have put their name to something. That is a materially different position from a publicist emailing a press release marked “EMBARGOED UNTIL 9AM” to a list of people who never asked for it.

The second support was editorial rather than journalistic. The Ingelfinger rule, set out by Franz J. Ingelfinger in a 1969 editorial in the New England Journal of Medicine, held that the journal would decline to publish findings already disclosed elsewhere. A near-identical policy had been stated by Samuel Goudsmit at Physical Review Letters in 1960. The effect was to give researchers a powerful reason not to talk to reporters early: doing so could cost them the publication. For decades, the embargo system rested at least as much on that threat to the author as on any agreement with the press.

That support has weakened. Preprint servers — arXiv, bioRxiv and others — made it normal for findings to be public, citable and covered long before the peer-reviewed version appeared, and many journals have loosened their Ingelfinger requirements accordingly. An embargo on a paper whose preprint has been online for six months protects very little.

Outside science, the honouring of embargoes is closer to a courtesy. It persists where the relationship matters more than the individual story, which is much of the beat reporting described in how editors decide what to cover, and it evaporates where it does not.

Documented breaks, and what followed

Most of what circulates about embargo breaks is folklore: everybody has heard that an outlet was blacklisted, and nobody can name it. There is one reasonably systematic public record, and it is worth using. Embargo Watch, run by Ivan Oransky and publishing since February 2010, has logged embargo breaks and the responses to them for over a decade. Below are cases from that record, with what actually happened next.

The Reuters case is the most instructive because it is the exception. In March 2019 Reuters broke an embargo on two studies being presented at the American College of Cardiology's annual meeting in New Orleans and published in the New England Journal of Medicine. Reuters described the break as inadvertent. The ACC revoked Reuters' press access to its 2020 annual meeting in Chicago and said it would no longer supply the wire with embargoed material ahead of future meetings. The NEJM, whose papers they were, imposed no sanction at all.

Two things follow from that split. The first is that the sanction came from the body controlling physical and informational access — a conference press room — rather than from the publisher. The second is that even a wire service of Reuters' standing can be excluded, which tells you the norm has some force where the gate is real. Embargo Watch's own assessment of such sanctions was that they “seem fairly rare”, and noted they may happen more often than organisations announce.

The other cases show the usual response, which is not punishment but surrender. When the UK Press Association ran a wire story ahead of schedule on a small endometriosis drug study in December 2019 — research involving sixteen women and mice, promoted by a University of Edinburgh press release — PNAS lifted the embargo early and said: “A wire story from the UK Press Association was published ahead of the scheduled embargo. We are investigating the circumstances.” No sanction on the Press Association is recorded. In June 2020 Science lifted an embargo on a dog-breeding study immediately after a break. In April 2021 the journal publishing a mantis shrimp larvae study did the same after The Guardian went early.

That is the pattern worth carrying: an embargo break is treated as an event to be managed, not an offence to be punished. The moment one outlet publishes, holding everyone else to the embargo only penalises the compliant, so the embargo goes. The breaker keeps the traffic from being first.

An embargo is not an exclusive

These get conflated constantly, including by people offering them, and they solve opposite problems.

An embargo distributes the same material to many outlets and fixes a single release time. Everyone publishes at once. Nobody is first, so nobody gains from being first, which is precisely what removes the incentive to rush and is the whole justification for the practice. What a source buys with an embargo is breadth and simultaneity — a wall of coverage at one moment — plus the reporting quality that comes from journalists having had time to check.

An exclusive gives the material to one outlet and to nobody else. That outlet has a scoop and every incentive to give the story prominence, because it owns something its competitors do not. What a source buys is depth and placement in one venue, at the cost of everyone else, some of whom will decline to cover a story they were passed over on. Timing is usually agreed too, which is why the two are confused, but the scarcity is the point rather than the clock.

The combination — an exclusive held under embargo until a set time, with a wider release following — is common and is a different offer again: one outlet gets to be first by a margin, the rest get the material afterwards. Anyone negotiating this should be explicit about which of the three they mean, because there is no authoritative industry definition of any of them. These are conventions of practice, not standards, and usage varies between countries, beats and individual desks.

One practical asymmetry is worth stating. An embargo is a constraint you are asking a journalist to accept for your benefit. An exclusive is something you are giving them. That is a large part of why exclusives are honoured more reliably than embargoes: the recipient's interest and the source's interest point the same way.

Is an embargo enforceable at all?

In ordinary practice, no — and where something like enforceability exists, it is the agreement doing the work rather than the embargo.

Start with the commonest case. A publicist emails a release marked “embargoed until Tuesday 9am” to a list of journalists who did not request it and did not agree to anything. Nothing was negotiated and nothing was accepted. A journalist in that position has received information and attached to it a unilateral instruction from a stranger, which is not how obligations are created. Many journalists will honour it anyway, out of habit or because the relationship is worth more than the hours; some publicly refuse unsolicited embargoes as a matter of policy. Neither is breaking a rule, because there is no rule.

The position changes where access was conditional and the condition was accepted. EurekAlert!'s digitally signed embargo agreement is the clearest example, and conference press-room registration and formal lock-ups work the same way: you got the material because you agreed to terms, and you agreed before you got it. That is contract-shaped. We could not locate a publicly documented case of a press embargo being litigated as a breach of contract, and we are not going to imply that one exists; the remedy used in practice is withdrawal of access, which is faster, cheaper and sufficient.

Then there is a third category that is often mistaken for embargo enforcement. Some pre-release information is governed by law independently: market-sensitive corporate and economic data, information under court reporting restrictions, material covered by official secrecy provisions. Publishing early there can attract real legal consequences — but the law is doing that, not the embargo. Confusing the two leads people to believe embargoes have a general force they do not have.

So the honest answer is that an embargo is enforced by access, reputation and repeat business, and works in exact proportion to how much the recipient needs the next thing you have. A science desk that wants the next press package is well controlled. A tech blog covering your funding round, which will never need anything from you again, is not controlled at all. That is the same asymmetry that governs whether anything gets read in the first place, as we found in whether press releases still get read and in how many pitches a journalist gets.

Methodology and limitations

The definition of an embargo and the account of signed embargo agreements are from EurekAlert!'s published help and FAQ pages, read on 24 September 2026; EurekAlert! is operated by the American Association for the Advancement of Science and is the principal distribution channel for embargoed science news, so it is a participant in the system it describes. The documented break cases are drawn from Embargo Watch, a specialist blog run by Ivan Oransky publishing since February 2010, with the Reuters and PNAS cases read in full from their original posts and the remainder taken from the blog's own index of cases. The Ingelfinger rule account, including the 1969 NEJM editorial and the earlier 1960 Physical Review Letters position, is taken from the encyclopaedic summary linked below rather than from the original editorials, which are behind publisher access; the primary citation is given so a reader can go to it. Four limits apply. First, there is no complete register of embargo breaks anywhere; Embargo Watch is one person's log, it covers science and medical journalism far more thoroughly than any other field, and organisations have every reason not to publicise either breaks or sanctions. Second, that skew means this guide describes the corner of journalism where embargoes are strongest, and the picture in technology, consumer and corporate reporting is likely looser than the cases here suggest. Third, several publications' own embargo policies could not be read: pages at Nature, the New England Journal of Medicine, the BMJ, Science and the BBC's editorial guidelines were behind authentication or refused automated access, so none of them is quoted. Fourth, we found no legal case law on press embargoes and make no claim about how a court would treat one.

Conclusion

Embargoes hold where they are worth something to the person being asked to hold them, and nowhere else. The system is strongest in science and medical journalism because access there is gated, signed for and needed again next month; it is weakest wherever a publicist emails a stranger an instruction. The documented record shows what happens when it fails, and it is not dramatic: the embargo is lifted for everyone, the story runs, and the outlet that went early is generally not punished. The single documented sanction in our sample came from a conference press office rather than a journal, which tells you where the power sits. For anyone planning an announcement, the practical reading is that an embargo is a coordination device rather than a protection, that it is worth asking for only from people who need something from you again, and that if the timing genuinely cannot slip, the answer is not to give the information out early at all.

Frequently Asked Questions

Sources

1. policy document EurekAlert!, American Association for the Advancement of Science. Help and FAQ. Definition of a news embargo; requirement that reporters and public information officers digitally sign an embargo agreement. Read 24 September 2026.
2. specialist reporting Oransky, I., 1 April 2019. Reuters banned from cardiology meeting after embargo break, Embargo Watch. American College of Cardiology 2019 meeting; access to the 2020 meeting revoked; no NEJM sanction.
3. specialist reporting Oransky, I., 2 December 2019. UK Press Association breaks embargo on tiny study of potential endometriosis drug, Embargo Watch. Includes the PNAS statement quoted above.
4. specialist reporting Oransky, I. Embargo Watch, running since February 2010. Source of the case index, including the June 2020 Science dog-breeding break, the April 2021 mantis shrimp break and the August 2018 Fields Medal timezone break.
5. academic Ingelfinger, F. J., 1969. Definition of “Sole Contribution”, New England Journal of Medicine 281(12), 676–677. The origin of the Ingelfinger rule. Cited from the secondary account below; the original is behind publisher access.
6. reference work Ingelfinger rule and News embargo, Wikipedia. Used for the history of the rule, the 1960 Physical Review Letters precedent and the effect of preprint servers. Tertiary source, used only where the primary document could not be opened.