FZ1073: Why Some Rumors Died in a Day and Others Outlived the Facts

Five days of online conversation after the attack on the Dubai to Tel Aviv flight

The rumors that survived were the ones that took a side

On 30 September, the co-pilot of flydubai flight FZ1073 attacked the captain on the way from Dubai to Tel Aviv. Passengers and crew stopped him, and two off-duty pilots landed the aircraft safely in Saudi Arabia. Online, it took days for that simple account to emerge. For many people, it never did.

Vinesight followed about 51,000 public posts in more than a dozen languages over the five days after the attack. One pattern explains most of what we saw. False details with no political value disappeared as soon as someone corrected them. False claims that gave credit or blame to a country survived the correction, and some reached most of their audience after it.

That pattern matters well beyond aviation. The Dubai to Tel Aviv route is one of the most visible results of normalization between Israel and the UAE. The claims that lasted pointed at Iran, Pakistan, Israel and the UAE itself, and each one tied the attack to an older conflict. For anyone who wants to understand how a single incident can strain relations between states, this is the mechanism to watch. The interactive charts below let you follow each claim hour by hour.

The first day decided what most people saw

41% of all engagement over the five days came on the first day. On that day, nobody outside the investigation knew who the crew were, which weapon was used, or why. The first false claims appeared within an hour of the distress call, before any official account existed.

For a reader, the practical meaning is uncomfortable. If you followed this story, you most likely followed it on day one, and the version you remember may be one that was later corrected.

A correction only works if the same people see it

The clearest example is the claim that the two pilots were a Russian and a Ukrainian who fought in the cockpit. It appeared 55 minutes after the distress call, attributed to Israeli television. Israeli outlets refuted it about two and a half hours later and reported that the attacker was Omani.

The refutation changed what people wrote, but it did little to change what people saw. After it, 60% of the posts about the claim corrected or questioned it. Those posts received 12% of the claim's engagement, while the posts that repeated it received 71%. In total, 85% of the claim's engagement came after it had been refuted.

Part of the reason is language. 87% of the claim's engagement came from posts not written in English. The biggest single post was a Spanish-language news video, published four and a half hours after the refutation, that still reported the first version. That one video produced half of the claim's engagement. Two days later, a Russian outlet reported that the nationalities had started as a joke in a chat group of Israeli journalists. That explanation reached 0.6% of the audience the claim had already found.

The lesson is simple. A correction that exists is different from a correction that arrives, and most of the people who saw this rumor never saw the denial.

People accept corrections that cost them nothing

Not every correction failed. On the first day, 97% of the posts that named a weapon said it was a knife. On 3 October, UAE prosecutors said that the co-pilot had used the cockpit crash axe. Within a day, 87% of weapon mentions had switched to the axe.

The weapon had no political value. No country, community or cause gained anything from the knife, so nobody defended it. Nationality, motive and responsibility were different. Every answer helped one side and hurt another, and those claims held.

This gives readers a useful test for any breaking story: ask who benefits if a claim is true. When a claim hands credit or blame to a country or a group, it is the kind of claim least likely to be corrected in front of you.

When the facts settled, the blame started

As the basic facts became clear, the conversation changed subject. On the first day, errors about what happened made up 19% of posts. By 3 October, they were under 5%. Over the same period, claims about who was responsible grew from 3% of posts to 13% on 2 October, and they stayed near 8% afterwards.

Each of these claims connected the attack to an existing rivalry. Claims that Iran directed it circulated from the first morning and peaked after President Trump said on 1 October that the co-pilot appeared linked to Iran “based on what I'm hearing”. No evidence has been published so far — though the alleged link has also not been denied, several reports suggest the pilot may have acted as a lone-wolf and was inspired by jihadist ideology. Claims that tied the attacker to Pakistan peaked on 3 October, after the co-pilot had already been identified as an Omani national. Of the Pakistan posts published after that identification, 95% did not mention it. A smaller set of posts pushed a conspiracy theory that UAE officials were complicit.

The conspiracy theory that Israel staged the attack grew from about 2% of daily posts to 6%, and they stayed there while other errors faded. More than 1,300 accounts published them. One in seven used the same piece of evidence: a photograph of Prime Minister Netanyahu with a passenger still in a shirt with blood on it. An image released to show national unity became the main exhibit for the opposite story.

Every audience found its own hero

Credit for the rescue followed national lines too. Three quarters of the posts in Indian languages mentioned the Indian captain, Smit Machchhar, and only 2% mentioned the Israeli passengers. His share of all engagement grew from 14% on the first day to close to 60% by 2 October. The Israeli passengers who entered the cockpit had their peak on 1 October, at 19% of the day's engagement, and then faded.

The two off-duty pilots who actually landed the aircraft received 3% of the engagement over five days. That is about the same as the false claim that an Israeli passenger flew the plane. Their names were not released, and no national audience adopted them. Each audience told the version in which its own people saved the flight.

What this means

Identity tends to explain when a rumor sticks, even after corrections. Any claim that tends to validate our belief or identity will likely be amplified and survive deeper examination.