PHOTO: AI is reshaping how information warfare is fought, denflinkegrafiker on Pixabay
On 1 March 2026, Iran’s Revolutionary Guards Navy suggested it had struck the USS Abraham Lincoln, possibly sinking the US aircraft carrier. The claim set off a flood of videos showing a ship in flames and victory being celebrated. The carrier was undamaged. The footage was fake. Yet the victory lived on, as the correction never caught up.
This was not an isolated incident. A deluge of artificial intelligence (AI)-generated videos and images has overwhelmed the internet. Videos showing huge explosions that never happened, soldiers protesting when they were never there or military vessels destroyed when they were never harmed have introduced a new layer of uncertainty into modern warfare. The public are left to question whether each image they come across is real or fake before they can allow themselves to react. As the line between authentic and synthetic evidence becomes increasingly blurred, the consequences extend far beyond simply being deceived, threatening how the public trusts and perceives war.
AI has accelerated this shift. In the hands of bad-faith actors, manipulated content can now be produced and deployed with unprecedented speed, scale and sophistication. AI is no longer just a tool for creating content. It has become a tool of war.
The battle off the field
Information manipulation and propaganda have always existed alongside war. AI has simply changed the economics. Paired with social media, AI has amplified this misinformation on a far greater scale. It makes information warfare much more powerful and effective, offering both operational and strategic advantages that were previously out of reach.
Early in the Israel-Hamas conflict, misinformation relied more on recycled or miscaptioned real footage. Now, with easy access to generative AI, almost anyone can produce realistic simulations of warfare at little to no cost. More than 110 AI-generated images and videos circulated in the first two weeks of the war between the US and Iran, The New York Times reported.
Creation is only half of it. AI can generate persuasive text, images and videos in seconds, while AI-driven bots share, like and comment on these posts to create the illusion of widespread support. Misinformation then becomes a powerful weapon in information warfare, capable of influencing public opinion, sowing discord and deepening divides.
AI as a Weapon of War
As technology advances and data increasingly becomes a critical asset in conflict, AI is positioned to become a national strategic resource. How it is used, however, will determine whether it protects the truth or drives a structural collapse of trust in authentic documentation. The latter is winning out. It takes many forms, from trolling and memes to fake satellite imagery and manipulated battlefield footage.
The US-Iran war is a key example, with each state taking a distinct approach to shaping its narrative. The official White House account on X posted a video, made without AI, that spliced footage of US strikes with clips from John Wick and Top Gun. Captioned “Justice the American Way”, the video casts US strikes as justice being served and presents the war as a victory in progress, a message of dominance aimed at audiences at home.
Iran, meanwhile, relied heavily on AI-generated media for very similar purposes of asserting military strength and exaggerating the damage inflicted. In one instance, the Tehran Times, an Iranian English-language newspaper, shared an AI-generated satellite image on X purporting to show an American radar destroyed by an Iranian drone strike. The post attracted almost one million views, despite it being a fake.
Such misleading AI-generated content erodes trust in authentic documentation, as the information environment becomes crowded with manipulated evidence that must compete with authentic coverage.
Iran is not alone in this. In Ukraine, Russia applies the same logic but to the front itself.
Between June and July 2026, Russian sources posted multiple AI-generated flag raising videos over areas of Ukraine that had neither been seized or infiltrated. The village was real. The street was real. What gave the videos away was the improbable proportions of the Russian soldiers and flags. Failure to capture land, to meet Kremlin-set deadlines or to shape a narrative of likely victory has led back to the same use of AI generated content, as assessed by the Institute for the Study of War.
These acts can no longer be seen as an online nuisance. The use of AI as a weapon now feature in the information warfare playbooks of actors across Russia, Mali and China.
Targeting Democracies
The reach of AI in information warfare no longer ends on the battlefield. States are now turning the same tools on democracies, and on the elections and public debates.
The Kremlin spends around £1.3 billion (AU$2.46 billion) each year on manipulating information, according to British Prime Minister Burnham in his address to the UN General Assembly. Russian agencies have used bots, fake websites and falsified articles, and these tactics could become more dangerous when combined with AI.
This is no longer speculation. In May 2026, leaked documents from Russia’s Social Design Agency (SDA) revealed plans to use AI to distort state decision-making and sway elections in Europe. A file titled “Projects 2026” sets out a database to monitor nearly 10,000 social media accounts of influential Europeans, and an “AI News” project to create more than 500 short videos for French audiences across multiple platforms.
This is not a new strategy. This replicates the same logic and model used by China’s state-aligned company GoLaxy, which was reportedly piloted against the US in 2025. Pro-Iranian networks have used the same tactic, pushing AI-generated videos tailored to different audiences that drew millions of views. These actors are no longer broadcasting propaganda. They are studying audiences, identifying their beliefs and vulnerabilities, and tailoring content.
When Belief is No Longer the Point
Whether any of this works is harder to measure than how it spreads. Some of the most effective content is not meant to be believed at all.
Take Iran’s AI-generated Lego themed videos, which use humour and parody to mock Donald Trump and US foreign policy, depicting the conflict as an exaggerated spectacle. The videos span over several minutes and at times contain factual inaccuracies, yet their garish visuals keep audiences watching. That may be precisely the point.
Propaganda does not necessarily need to be believed to have an effect. Even when viewers recognise what they are watching is fake, the narrative behind it can keep circulating. For Iran, these AI-generated videos are of particular importance. Given Iran’s conventional military disadvantage against the combined forces of the US and Israel, undermining confidence in the US’ military objectives may be a fundamental way for Tehran to gain ascendancy beyond the battlefield.
Propaganda also affects people differently. People already dissatisfied with their circumstances are particularly vulnerable to the foreign interreference online, said British Prime Minister Burnham. AI tailored influence campaigns can home in on vulnerabilities that already exist, reinforcing narratives of division and distrust.
The measure of success, then, is no longer dependent on whether the propaganda was believed. The danger is that it may win attention and leave audiences more uncertain than before.
PHOTO: The influence of propaganda, media and money on the human mind, HazarMan on Pixabay
The Cost of Doubt
The deeper damage is trust. Once fakes are everywhere, genuine footage risks being dismissed as fake. As AI makes synthetic evidence more convincing, authentic coverage becomes easier to wave away. This is known as the “liar’s dividend”, where the exposure of even one single fabricated image can be used to undermine the trust in genuine evidence as well.
The danger is greatest when fakes are disguised as credible. Russia’s “Projects 2026” calls for fictitious research institutes that give Kremlin narratives the appearance of legitimate sources. This strategy is already underway. A site called the “World Centre for Strategic Studies”, has published unsigned analysis aimed at Western audiences. Once these fabricated sources are cited by a legitimate outlet and gains credibility online, it can then be picked up and reproduced by AI systems. A manufactured narrative can then begin to look like an established fact.
Defences are struggling to keep up. X now suspends accounts that post unlabelled AI-generated content depicting armed conflict from receiving revenue for 90 days. But this penalty may do little, as many of these accounts including Pro-Iranian networks have appeared focused on spreading their narratives as opposed to making money.
The result is a battlefield of doubt, where the truth has become easier to deny.
The Test Ahead
The USS Abraham Lincoln was never sunk, yet the footage of it burning travelled further than any correction. It is tempting to treat that as a problem for the future, but the threat is closer than it seems. Just as propaganda does not need to be believed, AI does not need to win the argument. It only needs to reach the people before the facts do and leave behind a doubt that no correction can undo.
The next fake will come. So will the next truth. Whether either is believed will be the real test.
