The Strait of Hormuz is thousands of kilometres from Washington, but Donald Trump has nonetheless declared it ‘American territory’.
His recent claim followed more military action against Iran and a US blockade of the strategic waterway, and was accompanied by the US president sharing maps on Truth Social depicting the strait as belonging to the US and in its control, while also claiming billions of barrels a day in oil were being shipped under the watch of US warships.
Iranian officials quickly rejected the claim, insisting the waterway “will remain Iranian”, and live maritime shipping data available online clearly showed the only boats transiting the Strait were under the watch of Iran.
While some claims are easier than others to discount, competing versions are part of a larger phenomenon: the battle for narrative control.
And, in today’s military conflicts it is the internet and its easy deluge of mis- and disinformation that is increasingly redefining the fog of war.
First coined by Prussian military strategist Carl von Clausewitz in his 1832 book On War, the ‘fog of war’ describes the uncertainty of information on the battlefield, which can lead to mistakes, miscalculations and friendly fire.
But in the post-truth era, that fog is spreading, and more complicated than ever to make sense of.
Synthetic media enables the plausible deniability of whether it has happened or not. States and actors that are involved in conflict can now dismiss real evidence as fake.
Dr Ayesha Jehangir, UNSW
The ‘digital’ fog of war might better describe Clausewitz’s concept as we know it today, in order to better understand the chaotic information environment surrounding conflict, where authentic reporting, misinformation, propaganda and manipulated media compete, to shape how audiences understand events.
For governments, institutions, and bad actors alike, the rise of social media and artificial intelligence has enabled something new: the strategic manufacture of doubt.
In modern war, seeing is no longer believing.
Dr Ayesha Jehangir, a journalism academic at UNSW who researches how conflict and human suffering are represented in media, says the modern fog of war is no longer simply about a lack of information, but an overload of competing narratives online.
“Our real problem is not detecting what is fake. AI is going to create this monster for us, where we are going to struggle to detect what is true, because we are not going to trust anything,” she says.
“The use of anything fake, like propaganda for instance, has been part of warfare for a long time… synthetic media enables the plausible deniability of whether it has happened or not. States and actors that are involved in conflict can now dismiss real evidence as fake.
“That is an extension of what scholars have called fog of war, for a very, very long time.”
In 2025, the Oslo Peace Research Institute recorded 65 state-based conflicts across 35 countries – the highest number since records began in 1946.
And there continue to be major developments in both the scale of these conflicts worldwide, as well as the capacity for news manipulation. The Reuters Digital News Report 2025 found 58 per cent of people globally held concerns about the credibility of online news, worrying about distinguishing real news from fake. The greatest concern came from Nigeria (84 per cent), South Africa and the US (both 73 per cent).
With over half of the global population struggling to distinguish between reality and AI-manipulated falsehoods online, the content itself is now also shifting. Some estimates suggest over 70 per cent of social media posts are AI-generated as of March 2025, with a forecasted increase to 90 per cent throughout 2026, according to a report published by the Europol Innovation Lab, updated in 2024.
The consequences are already visible in conflicts where fabricated content circulates alongside genuine evidence, making even authentic documentation vulnerable to doubt.
Claims that a Palestinian child killed by Israeli strikes was actually a doll have been repeatedly debunked, but not before gaining significant traction online.
Videos of film make-up artists from 2017 were also recirculated as supposed evidence that injuries in Gaza were being staged amid the ongoing conflict.
In July 2026, AFP Fact Check reported on AI-generated footage purporting to show mass graves in Sudan, as fighting between the Sudanese army and the paramilitary Rapid Support Forces (RSF) continued. While the RSF and allied militia groups were accused, verification software found the video had a “99 per cent” likelihood of being AI-generated.
But the presence of fabricated footage can also misleadingly make consumers doubt real evidence of atrocities.
In October 2025, satellite imagery independently documented the aftermath of mass killings in El-Fasher. ABC News Verify geolocated footage of the aftermath to the same locations, while investigators identified networks of holes thought to be dug for the incineration and disposal of bodies.
The question is no longer simply: what is fake and being perceived as real? Increasingly, it is becoming: what is real and being dismissed as fake?
Researchers call it the ‘liar’s dividend’: where the existence of deepfakes and AI-generated content enables bad actors to dismiss authentic evidence as fabricated.
“The concept of the liars dividend… first appeared in Russia’s full-scale invasion of Ukraine… Particularly in 2022, when [Volodymyr] Zelenskyy deepfakes went viral, telling his troops to surrender,” says Jehangir.
A deepfake video of Ukraine’s Volodymyr Zelenskyy, left, and the real one on the right.
The result is a “cry wolf” effect, where genuine documentation is treated with the same suspicion as manipulated content.
Jehangir says that although many similar videos were quickly debunked due to the emergence of fact-checking as a dedicated unit within many media organisations, the existence of these videos demonstrates a broader problem: “Once synthetic media becomes ubiquitous, any piece of evidence that comes from any part of the world… can easily be contested.”
“AI-generated media are not just tools of deception, but also tools of doubt,” says Jehangir. “So, AI does not only end up producing falsehoods… It corrodes the credibility of truth itself.
“That’s why we refer to this time as a post-truth time.”
In such an environment, the fog has become so effective because of how easily it is being produced and circulated. Protecting what is real, rather than simply identifying what is fake, is more important than ever.
However, separating the authentic from the manufactured is becoming increasingly difficult. AI has made it possible to generate or alter convincing war footage, images, and narratives at scale, often faster than they can be independently verified or debunked.
UTS Centre for Media Transition researcher Michael Davis says the solution lies with a greater emphasis on verification practices.
“Over the last 40 years… people have been relying on photographic evidence to show what’s been happening in these conflicts,” he says.
“Now we have to be a bit more careful about what we find on social media and we have to make sure that we can corroborate it. The politicisation of news and information in conflicts is really where the main problem lies.”
We’re going to work out, over time, how much trust we think is reasonable in various contexts to give visual media.
Michael Davis, Centre for Media Transition
Despite existing alongside an overwhelming body of authentic documentation produced by journalists, aid organisations, and civilians on the ground, the playbook of deception has become increasingly apparent across social media, with fake bot accounts amplifying coordinated narratives, AI-generated images and videos, recycled footage, and manipulated media designed to muddy the waters.
In Myanmar, Facebook was used to spread false claims that Rohingya Muslims were plotting to overtake Buddhist communities, employing fabricated stories and inflammatory posts amplified by military-linked accounts in 2017.
Human rights investigators later identified social media as a significant driver of the hate speech that preceded violence against the Rohingya population. The aftermath continues today, with Rohingya refugees in the UK and US suing Meta for allegedly allowing anti-Rohingya hate speech to spread unchecked.
These deliberate dynamics continue to reveal themselves with state-based conflicts reaching a multi-decade peak. Among them is the US-Israel war against Iran, where competing actors have increasingly turned to social media, as a secondary battleground over how the conflict is represented and understood.
With virality peaking in March 2026, Iranian-linked outlets continue to circulate satirical ‘Lego-style’ memes and AI generated videos, that have arguably done more to humanise the country to western audiences than any media report.
Dr Jehangir says stylised representations of conflict can shape who audiences empathise with and how credibility is assigned online.
“They [AI images] are designed to provoke outrage, anger, but also empathy and solidarity,” she says.
“It produces suffering that is instantly consumable, you don’t have to be very close to it, even if you are a very distant spectator… It is shareable, and emotionally legible. It speaks to you in your emotional language.”
Using the US’s own pop culture to poke fun at Trump and the White House, Iran’s viral Lego social media output smartly reframes complex geopolitical tensions into digestible, emotionally engaging content, aimed at softening Iran’s hardline image, while drawing attention to the failings of its adversaries.
In February 2026, various claims surrounding the US missile strike on the Shajareh Tayyebeh girls’ school in Minab, which killed over 150 people, most of them primary school children, spread rapidly online. Viral posts on X falsely alleged Iran had admitted responsibility for the attack and shared a photograph purportedly showing an IRGC missile hitting the school.
There was no immediate evidence to support either claim.
President Trump told reporters he believed Iran was responsible for the strike, but four days later, The New York Times reported a US military investigation had concluded the school was struck by an American Tomahawk missile.
Trump continues to employ Truth Social and X to rapidly disseminate unsubstantiated claims about the conflict.
He also shares contentious AI-generated images and other unverified claims on a wide-scale, using synthetic media to reinforce a carefully constructed image of his own leadership.
President Donald Trump’s Truth Social posts on Canada and Iran.
For the public and journalists alike, the challenge is increasingly how to establish what can be trusted among competing claims and manipulated imagery.
Davis argues that concerns over emerging technologies are not new, with society historically adapting its expectations of new forms of media.
“The same kind of debates happened when photography arrived, and… even the printing press. People are always concerned about the threat of new technologies,” he says.
“We’re going to work out, over time, how much trust we think is reasonable in various contexts to give visual media.”
For Davis, the answer lies in established standards of verification.
“Journalism has very robust standards of verification built up over time, particularly over the last 50 years or so, for fact-checking information,” he adds. “That kind of epistemology of fact-checking still rests on those same kinds of journalistic standards.”
But within an evolving mediascape, the platforms distributing information at unprecedented speed also shape what audiences see.
The Communications Legislation Amendment (Combatting Misinformation and Disinformation) Bill 2024, which sought to strengthen transparency and accountability requirements for digital platforms, ultimately failed to gain sufficient support in the Senate.
I think it’s very important for us to slow down our consumption of war news. One of the most important things people can do… is to resist the pressure to react instantly.
Dr Ayesha Jehangir, UNSW
Davis says the proposed legislation had the right objectives, despite concerns around freedom of expression.
However, social media platforms themselves also have a significant role to play in determining what information reaches audiences at scale. For Davis, that responsibility extends beyond simply removing false information.
“There’s also ensuring the prominence of quality news and information,” he says. “So, instead of saying, we’re just a neutral host for whatever… what’s in the public interest for us to promote?”
Jehangir says the first step starts with audiences.
“I think it’s very important for us to slow down our consumption of war news,” she adds. “One of the most important things that people can do… is to resist the pressure to react instantly.
“Social media environments are designed to reward speed… You see, you feel, you share within seconds. But in conflict reporting, speed often comes at the expense of accuracy and context.
“So, being well-informed is important. But today, it is less about consuming more information. It’s about consuming it more deliberately.”
Main image montage of Gage Skidmore/Flickr photo/Trump Truth Social post and Iran Lego meme.
