Forty That Weren’t: Anatomy of an Atrocity Claim

The claim did not begin as a report. It began as a single television correspondent, embedded with the Israeli military, relaying what she said Israeli reserve soldiers had told her. As FactCheck.org’s review of the original broadcast clips documented, the correspondent attributed the count and the detail to separate remarks from a commander — in one clip citing “about 40 babies at least,” in another relaying that infants had had “their heads cut off, they said” — never claiming to have personally verified either detail herself. [1]

From that single, unverified, secondhand account, the claim moved through a very short chain. A spokesperson for Israeli Prime Minister Netanyahu told reporters that infants had been found “decapitated” in Kfar Aza — CNN’s own live coverage noted it could not independently verify the claim even as it reported the spokesperson’s statement. [2] Headlines and subsequent references quickly dropped that attribution and reported the claim as the event itself. Then, on October 11, at a White House roundtable with Jewish community leaders, President Biden stated without hedge that he had seen “confirmed pictures of terrorists beheading children.” [3] That evening, an administration official told NBC News that neither Biden nor the administration had actually seen any such pictures or independently confirmed the reports — his remarks, they said, referred to media and Israeli-official statements. [4] The Washington Post later reported something more damning still: that Biden had been advised by his own staff to cut the line specifically because the reports were unverified, and used it anyway. [8]

At no point in this chain did a second, independent party confirm the original claim. The head of state didn’t just repeat an unverified rumor — by the time he spoke, he had reportedly been told it was unverified, and presented it as something he had personally seen confirmed regardless. Each subsequent outlet’s citation was to the previous outlet’s coverage, not to new evidence. This is worth naming as its own technique, distinct from ordinary source hierarchy:

Pattern connection: this extends the platform’s existing source hierarchy technique — where IDF statements are treated as primary and Palestinian testimony as secondary — by showing what happens when even the Israeli institutional source (the IDF itself) never confirmed the claim, and outlets ran with it anyway, sourced only to a soldier’s remark relayed by a single reporter.

This is not a marginal effect. Decades of cognitive psychology research on the illusory truth effect show that repeated exposure to a claim increases how true it feels — regardless of whether the claim is actually true, and even when it contradicts what a person already knows. [14] A 2024 review in Current Opinion in Psychology found the effect holds for fabricated news headlines and conspiracy-style claims, not just neutral trivia, and that even statements people would otherwise recognize as implausible become more believable simply through repetition. [13]

The mechanism is “processing fluency”: a claim heard multiple times becomes easier for the brain to process, and the brain misreads that ease as a signal of accuracy — “this feels familiar, so it must be true.” Crucially, standard correction methods largely fail against it. Research published in Cognition found that labeling a claim as “disputed by fact-checkers” does not meaningfully reduce the illusory truth effect once repetition has done its work. [15]

This matters directly for this case: by the time the claim had been repeated by half a dozen major outlets and a head of state within 24–48 hours, no correction — however accurate or prominent — was psychologically capable of fully undoing the belief it had already produced in millions of people. The claim’s window of maximum repetition and its window of zero verification were the same window.

Mainstream wire and broadcast journalism has long-standing, publicly documented standards for exactly this situation — a shocking atrocity claim, from a single source, in a warzone, with high stakes for public opinion. At minimum, these standards require:

Independent corroboration before publication as fact. A single secondhand account — a soldier telling a reporter what another soldier told him — does not meet the bar for reporting a claim as confirmed. It meets the bar for reporting “an official said,” with that attribution preserved throughout every subsequent reference, not just the first one.

Proportional skepticism for extraordinary claims. The more shocking and mobilizing a claim, the more verification it requires before print — not less. A claim of mass infant beheading is about as extraordinary as war reporting gets.

Attribution that survives repetition. CNN’s initial coverage did technically attribute the claim to a spokesperson. [2] But once the claim was repeated across outlets and by officials, the hedge (“a spokesperson says”) was stripped away in practice, and headlines and public memory retained only the claim itself.

Correction proportional to amplification. A claim repeated globally and cited by a head of state requires a correction of matching scale and prominence when it collapses — not a quiet absence of further mentions.

Every one of these standards exists specifically to prevent what happened here. None of them functioned as a check in this case.

Within days, a striking number of the same outlets that had carried the claim explicitly said they could not confirm it — often quoting Israel’s own military declining to confirm its own government’s story:

CNN stated it could not independently verify the decapitation report, even as it continued reporting the Netanyahu spokesperson’s claim. [2]

NBC News quoted IDF spokesperson Maj. Doron Spielman directly declining to confirm it: “That specific report and that number I can’t confirm.” NBC stated it had not confirmed the reports itself. [4]

The Intercept reported that the IDF, asked directly, could not confirm the beheaded-babies claim at all. [5]

Haaretz, Israel’s own paper of record, went further than any Western outlet, concluding outright that some of the atrocity claims from that period — including this one — had been proven untrue. [6]

Al Jazeera reported the White House’s walkback of Biden’s remarks directly and prominently. [7]

It is also worth noting that Hamas itself denied the claim from the start. Hamas spokesperson Izzat al-Risheq told reporters the reports “spread lies about our Palestinian people and the resistance,” and PolitiFact reported him saying there was “no evidence to support such claims and lies.” [9] This denial does not settle the question on its own — an accused party denying an atrocity is unsurprising and not evidence of innocence — but it belongs in the record precisely because so much of the coverage treated the accusation as self-evidently true while treating the denial as barely worth a sentence.

In other words: the institution best positioned to confirm the claim — the Israeli military itself — never did. Multiple major outlets said so explicitly, in the same window the claim was still being repeated as fact elsewhere. That combination — simultaneous confirmation-seeking and confirmation-repeating within the same information ecosystem — is what makes the “quiet downgrade” so effective: the corrective reporting existed and was published, but never overtook the volume or prominence of the original claim.

What did not happen is the significant part. There was no equivalent wave of front-page corrections. No outlet that had run the original claim prominently followed up with equally prominent language stating the claim was unconfirmed or false. The claim simply stopped being repeated — it faded rather than being retracted. PolitiFact, to its credit, revised its original article rather than leaving an uncorrected claim live, but this is the exception, not the pattern. [10] [9]

This is distinct from the platform’s existing institutional silence technique, which describes outlets choosing not to cover an event at all. Here, outlets did cover it — loudly — and then quietly stopped, without ever performing the inverse of their original coverage. That gap deserves its own name:

This case follows a well-documented template, not a new one.

The Nayirah testimony, 1990. On October 10, 1990 — 33 years to the day before the beheaded-babies claim — a 15-year-old identified only as “Nayirah” testified before the US Congressional Human Rights Caucus that she had personally witnessed Iraqi soldiers pull babies from incubators in a Kuwaiti hospital and leave them to die on the floor. President George H.W. Bush cited the story repeatedly in building the case for the Gulf War. [17] It later emerged that “Nayirah” was the daughter of Kuwait’s ambassador to the US, that her testimony had been coached by the PR firm Hill & Knowlton as part of a multimillion-dollar campaign funded by the Kuwaiti royal family, and that investigators could not substantiate the claim of mass incubator killings at the scale described. [18] The structural parallel to 2023 is exact: a single emotionally devastating claim, from a source with an undisclosed interest, adopted without independent verification by media and a US president, used to build support for war, and never meaningfully corrected once its credibility collapsed.

The broader pattern. Media scholar Jack Shaheen spent over two decades cataloguing more than 1,000 Hollywood films and found a persistent pattern: Arabs and Muslims depicted overwhelmingly as violent, barbaric, and inherently threatening, with fewer than a handful of films offering positive or even neutral portrayals. [16] Shaheen’s research — and the academic field it helped establish — documents that this is not a series of isolated missteps but a durable cultural infrastructure that makes atrocity claims against Arab, Palestinian, and Muslim actors more readily believed, more quickly repeated, and less rigorously checked than equivalent claims would be against other groups. The “beheaded babies” claim did not have to fight against skepticism about Palestinian capacity for extreme violence — it moved through a media and public environment already primed to expect it.

Pattern omission, applied to the propaganda itself: reporting the 2023 claim in isolation — as most outlets did — obscures that this is at least the second time in 33 years an uncorroborated atrocity story about children, from a single interested party, has helped manufacture consent for a war against a majority-Arab or Muslim population. Naming that pattern is itself part of the correction the original coverage never provided.