Credible defence media and the fight against online misinformation

Earlier this year a claim began spreading online that a British warship had been struck by a Hezbollah missile in the eastern Mediterranean, initially appearing as speculation around an attack Hezbollah said it had carried out against an Israeli vessel, before changing fairly quickly into a much more specific story that the missile had actually hit a British ship.

HMS Dragon was the name that appeared most often online relating to this supposed incident. The Type 45 destroyer had been sent to the eastern Mediterranean as part of the British military presence protecting UK bases and interests around Cyprus, so to somebody seeing the claim without much other context it probably sounded plausible enough, and once screenshots began circulating the story started to acquire extra details as it went.

RFA Lyme Bay was mentioned too in some of the material being passed around. Within a fairl short period of time we were getting messages about it from readers and other people asking the obvious question: had a British warship actually just been hit by a missile?

We didn’t know immediately, and there was a period during which the only responsible answer was that we could not verify what was being claimed. Saying “we don’t know” isn’t a bad thing. So that is what we said publicly while we started checking it.

There is always pressure with stories like these moving at that speed to get something out quickly because everybody else is talking about it, particularly when screenshots are appearing from accounts which themselves appear to be quoting other sources, though at that point all we really had was a claim which had been repeated a lot, and repetition does not make something more reliable.

What we were doing behind the scenes

Several things were being checked at roughly the same time.

We were looking at the available open-source information to get a picture of where ships were, checking AIS information and other vessel tracking where it was available, and trying to work out whether anything we could see actually supported the claimed location of the incident.

A colleague was also watching aircraft movements around Cyprus and the eastern Mediterranean, because if a Royal Navy ship had genuinely suffered a serious missile strike then there was a reasonable chance that something else would start happening around it, perhaps helicopter movements, search and rescue activity, casualty evacuation or some other unusual pattern which might be visible through open-source tracking.

Military activity is not all visible on public tracking systems and warships quite routinely disappear from AIS, so the tracking could only ever form one part of the wider picture and we kept checking elsewhere.

We contacted the Ministry of Defence, UK Maritime Trade Operations and other sources as well, while continuing to look at what was circulating and, importantly, where some of the claims were actually coming from because by then people were quoting screenshots of screenshots and it was becoming difficult to tell which details belonged to the original allegation and which had simply been added afterwards.

That happens remarkably quickly with stories like this on social media.

One account says a British vessel may have been involved, another names HMS Dragon, somebody else adds that the ship has been damaged and within a few rounds of reposting the final version can look much more detailed and much more certain than the information it originally came from. People repeated it as fact because someone else repeated it.

Our checks weren’t producing anything that supported it, thankfully.

We couldn’t find evidence of the reported strike on a British vessel, we were seeing no activity that looked like the aftermath of a serious attack and the people we were speaking to weren’t corroborating the story either.

Then the MOD came back and confirmed that the claims were false. That answer came quickly, and in this particular case the speed mattered because the story was still moving online. We could go from saying that we had seen the reports and couldn’t verify them to telling people that the British government had directly denied that a Royal Navy vessel had been struck. We could shut the story down and answer the people reaching out to us.

The correction was therefore now travelling while the original story was still being shared, rather than arriving the following morning when thousands of accounts had already accepted the thing as established fact.

Why saying “we don’t know yet” matters

There is a wider lesson here about defence reporting, there is an extraordinary amount of defence information online now. Some of it is excellent, some of it comes from people with genuine expertise and some of it is anonymous material being repeated by accounts which have no obvious way of knowing whether the thing they have just posted is true.

For an outlet operating under normal editorial standards, there is a responsibility attached to putting something out under your own name. UK Defence Journal is regulated by IPSO, our journalists are named and we work under the Editors’ Code, and if we publish something significant which turns out to be wrong then we have to account for it.

At one stage the most useful thing we could tell people was that we knew about the story and were checking it, but we could not verify it. It is a fairly limited update, though it gives readers something solid to work with while the rumour is moving around them.

People were already sending the claim to us because they wanted somebody to check it. Once the MOD responded we could put an authoritative answer into the same information space, through an audience which was already paying attention to the story, and that meant the denial had a chance of catching up with the rumour before the rumour became accepted simply because everybody had seen it.

The MOD deserves some credit for the speed of the response here as well, that was really helpful here. Press offices quite reasonably cannot chase down every strange thing somebody posts online, and there are plenty of occasions where the government will either have no immediate answer or will be unable to provide one for operational reasons, but a claim that a deployed British destroyer has just been struck by a Hezbollah missile is obviously significant and in this case the answer came while it could still make a difference.

If it had taken until the next day, we’d still have done the same checking and we could still have reported that none of the available evidence supported the claim, though there would have been a much longer period where the information vacuum remained. That is the main thing myself and our team took from the whole episode.

The relationship between journalists and defence press offices gets built through loads of much less exciting exchanges, questions about deployments, equipment, exercises, timings, things which sometimes result in a story and sometimes don’t, and most of that work is fairly mundane. Then something like this happens and having people who know who to call, and somebody at the other end who understands why an answer is needed quickly, suddenly becomes much more important.

The Dragon claim was false.

For a few hours, though, the episode became a useful example of how this sort of thing now develops. A dramatic claim appeared, it was repeated internationally and picked up extra details, readers wanted to know whether it was true and we had to work through several different ways of checking it before we could give them a firm answer.

Open-source checks gave us part of the picture, sources filled in more of it and the MOD’s quick response allowed us to publish a firm denial while the claim was still circulating.

That is why the everyday relationship between defence organisations and the journalists covering them still matters.