Why Iran’s Lego Clips Are Anything but ‘Stupid’

While answering a question in July about the prospects for negotiations with Iran, U.S. Secretary of State Marco Rubio made a dismissive reference to an ongoing series of AI-generated Lego animation clips produced in Iran that mock U.S. and Israeli officials. Iran will be ready to do a deal soon, Rubio suggested, because “the price they are paying is very high, despite their tough talk and their flowery language and their stupid Lego videos they put out.” Rubio, speaking to reporters at a foreign ministers’ meeting in Manila, hadn’t even been asked about the propaganda campaign. That a sitting U.S. secretary of state felt compelled to address a cartoon campaign produced by a small group of individuals at an international security forum is itself revealing. The first of these videos to go viral appeared online in March. Most viewers assumed it was a joke. A Lego-figurine Donald Trump jolts awake in the Oval Office as Iranian missiles rain down on a U.S. warship. Within days, it had been viewed millions of times. The video was produced by a group called Akhbar Enfejari (“Explosive News”), a small outlet operating from inside Iran reportedly with a team of roughly 10, using consumer-grade hardware. Over the following months, the group released more than a hundred videos, distributed across X, YouTube, Instagram, and Telegram, along with audio tracks on Spotify. YouTube suspended the group’s channel in April. The content kept circulating anyway. While answering a question in July about the prospects for negotiations with Iran, U.S. Secretary of State Marco Rubio made a dismissive reference to an ongoing series of AI-generated Lego animation clips produced in Iran that mock U.S. and Israeli officials. Iran will be ready to do a deal soon, Rubio suggested, because “the price they are paying is very high, despite their tough talk and their flowery language and their stupid Lego videos they put out.” Rubio, speaking to reporters at a foreign ministers’ meeting in Manila, hadn’t even been asked about the propaganda campaign. That a sitting U.S. secretary of state felt compelled to address a cartoon campaign produced by a small group of individuals at an international security forum is itself revealing. The first of these videos to go viral appeared online in March. Most viewers assumed it was a joke. A Lego-figurine Donald Trump jolts awake in the Oval Office as Iranian missiles rain down on a U.S. warship. Within days, it had been viewed millions of times. The video was produced by a group called Akhbar Enfejari (“Explosive News”), a small outlet operating from inside Iran reportedly with a team of roughly 10, using consumer-grade hardware. Over the following months, the group released more than a hundred videos, distributed across X, YouTube, Instagram, and Telegram, along with audio tracks on Spotify. YouTube suspended the group’s channel in April. The content kept circulating anyway. The standard response would be to treat Akhbar Enfejari as an oddity, more prank than policy threat. That is a mistake. The campaign is proof of concept for a new operational model of propaganda, one that generative AI has made possible. States have long sought to mobilize supporters and attack adversaries. What generative AI changes is the organizational threshold. It reduces the size and composition of the staff needed to create this content—you do not need separate teams of writers, translators, graphic designers, animators, or distributions staff. These tasks can be done by a very small team, like Akhbar Enfejari’s. It also reduces the amount of time needed to produce content. Polished content can be created in a few hours. Finally, generative AI does not require staff to have sophisticated technical skills. Crucially, generative AI also creates the ability to produce multiple and different streams of influences at once. One small team—in this case, Akhbar Enfejari—rapidly produces culturally distinct messages for two different audiences at a speed and volume that once required a much larger organizational scale and resources. English-language satire circulates among audience resistant to Iranian messaging—the “stupid Lego videos”—while Persian-language videos reinforce Iran’s civilizational continuity, frame attacks on Lebanese civilians as a shared struggle, and generally reinforce solidarity among audiences already receptive to Iran’s ideological and religious vocabulary. The English-language Lego-style videos, by contrast, are not seeking to persuade. They do not need to be believed. Instead, they are watched, laughed at, and shared. They capture attention, including from adversaries, and that attention feeds on itself. Rubio’s unsolicited reference, for instance, was followed within hours by a new clip portraying him as “Little Marco,” illustrating the campaign’s feedback loop. Akhbar Enfejari’s campaign has penetrated the highest levels of American discourse without even requiring officials to engage with its claims. For most of the past century, high-quality propaganda required infrastructure that left a traceable footprint. Russia’s Internet Research Agency, before it was exposed, employed hundreds of staff in a dedicated St. Petersburg office building. The Islamic State’s Al-Hayat Media Center required physical production facilities and a permanent presence in an active war zone. The infrastructure required by such operations let investigators trace campaigns back to their sources. Content that reads as obviously state-sponsored is discounted by exactly the audiences it most needs to reach, a reliability penalty documented in communications research since the 1950s. That is why the most effective covert operations—from Soviet active measures to China’s state social media network—worked through concealment rather than open messaging. These operations nonetheless faced a persistent trade-off between visibility and credibility. Thus, states had to choose between polished official messaging that was ignored and deniable, grassroots-looking content that was expensive and risky to produce. Generative AI eliminates such trade-offs. A review of a sample of Akhbar Enfejari videos shows two distinct product lines running in parallel on the same channel, an arrangement that earlier propaganda operations could never have afforded to sustain. The English-language videos, which make up the majority of the group’s video productions, follow a deliberate escalatory arc. A recurring Lego Trump and senior U.S. officials appear first in domestic decadence, then moral corruption (referencing Jeffrey Epstein and Illuminati iconography), then physical degradation, and finally civilizational collapse as Trump emerges from a pile of skulls. The videos are precise about geography and military hardware. They depict the Strait of Hormuz, label the Burj Al Arab (a luxury hotel in Dubai that defines Emirati wealth), and name missile systems. The conspiracy references are not incidental; they are precise insertions into narratives already circulating in Western right-wing and anti-establishment communities conveying, essentially, “We know what you know, and we’re saying it out loud.” The videos do not need viewers to believe any specific claim. Research on disinformation exposure shows institutional trust erodes even when the underlying claims are rejected. Corrosion, not conversion, is the aim. A quick scan of social media shows that the format has already escaped its source. Accounts with thousands of followers as well as ordinary individual users are either sharing Akhbar Enfejari’s videos directly or producing their own Lego-style clips to criticize U.S. foreign policy. Not so incidentally, the Lego brand is doing a lot of work here. It is globally recognized, aesthetically inoffensive, and carries no political charge of its own—exactly the qualities the disruption track needs. Rendering violent scenes of missile strikes and burning buildings in Lego bricks reads as satire rather than incitement, which is why it slipped past content moderation systems designed to catch graphic or hateful content. (YouTube suspended the Akhbar Enfejari channel under its spam and deceptive practices policy rather than its violent content rules.) This format also fits into an existing internet genre. Lego parody and combat animation, known as brickfilms, have circulated on YouTube for decades unconnected to any state or ideology; Akhbar Enfejari’s videos could pass as an entry in a recognizable tradition rather than propaganda. Running simultaneously on the same accounts, in Persian, is a functionally different line of videos featuring traditional Noha and Latmiya (genres of Shiite Muslim devotional lamentation) mourning chants, invocations of Imam Ali, Shiite historical references (specifically the Battle of Karbala), and references to Iran’s “Sacred Defense” framing of the 1980-88 Iran-Iraq War. Designed for domestic Iranian and Persian-speaking regional audiences, this material adopts none of the ironic distance of the English-language videos. Rendering Karbala or wartime sacrifice in the mocking Lego style of the English track could trivialize the sacred history from which the message derives its legitimacy. Instead, the Persian track depends on devotional sincerity, history, resistance, and highlighting civilians suffering in Lebanon and the students killed in Minab. The scale achieved is difficult to overstate relative to the resources involved. Two coordinated pro-Iran networks generated more than a billion views on X in the first month of the current war, with Iranian and Russian diplomatic accounts driving the bulk of that amplification. Akhbar Enfejari approached that reach with a handful of people and consumer hardware, while Iran’s foreign ministry preserved public deniability even as the group itself admitted that the Iranian government was one of its clients. The format has already begun to travel—including to the government it was built to needle. Trump has continuously posted AI-generated content on his social media accounts to send messages not only to his audiences but to opponents and adversaries as well. Recently, he posted an AI-generated war video depicting a fictional strike on Iran’s Kharg Island. U.S. Vice President J.D. Vance did not defend the video as accurate. He said Trump was “sending a message to the Iranians.” That same logic is what makes the Lego videos work: not truth, just visibility. The country that called this tactic stupid is adopting it. What should concern policymakers is not this campaign specifically but how replicable its division of labor is. The Akhbar Enfejari model shows that a small, cheap, plausibly deniable team can now produce propaganda that matches the reach of a state media apparatus, widening circulation among outsiders while maintaining ideologically disciplined content for insiders. Current policy on platform enforcement and labeling is not equipped to face this new model of propaganda. Platform enforcement targets individual channels, but Akhbar Enfejari’s content outlived its own deplatforming because the real distribution engine was a network of amplifiers, state media accounts, diplomatic handles, and sympathizer reposts that no single takedown touched. Several videos embed QR codes directing viewers to affiliate channels, allowing reuploaded copies to retain pathways back to the campaign. YouTube suspended Akhbar Enfejari’s channel in April, but the suspension did not stop the videos from spreading and being reposted elsewhere. The platform has since lifted the ban, and Akhbar Enfejari is live on YouTube again with accounts also on TikTok, SoundCloud, Rumble, UpScrolled, Threads, Pinterest, and Bluesky. AI content labeling assumes that audiences discount labeled material. The evidence instead shows that labeling does not reliably reduce sharing once content has already found emotional traction. Moreover, this model produces many versions from many nodes, none of them singularly authoritative, all of them disposable, and now demonstrably imitable by actors with no connection to the original campaign. Closing the gap starts with treating this as a genre, not an incident, using the same tools built to track militant propaganda, such as narrative coding, audience targeting, and longitudinal comparison across releases, applied systematically rather than fact-checked once and forgotten. That same discipline should extend to distribution. The goal should be mapping the amplification networks—the state-linked handles and sympathizer nodes—that give fringe content its real reach. The harder shift is analytical. Countermessaging keeps treating this content as an argument to rebut, when its appeal is the audience relationship it builds, the sense of being let in on something, the ironic distance that makes sharing feel like commentary rather than endorsement. Akhbar Enfejari has already been deplatformed once and kept going. The model it pioneered is not contained. It has already been copied by actors with no stake in Iran’s war, and it will not stay confined to the conflict that produced it.