AI-Generated Lego Propaganda: What Explosive Media's Viral Operation Reveals About Synthetic Influence at Scale
Explosive Media's AI-generated Lego videos hit millions of views. What their production pipeline means for architects and BIM professionals using synthetic imagery.
The Pipeline, Not the Politics
In February 2025, a pro-Iranian group calling itself Explosive Media posted its first Lego-animated political video. The original YouTube channel had averaged a few hundred views per clip. Within weeks of the format pivot, individual videos were reaching millions of views on X, Telegram, and TikTok — scripted, produced, and edited entirely via AI tools, according to reporting by WIRED (cited in Ars Technica’s coverage). More than a dozen videos have been released since. That is not a content strategy. That is a production pipeline.
←TODAY: A small anonymous team produces geopolitical AI animation at scale in 2025, timed to breaking news cycles within hours.
→3012: Synthetic media pipelines are standard infrastructure for any actor — state, corporate, civic — willing to invest twelve months in workflow design.
Fulcrum: The production speed is the weapon, not the aesthetics.
The mechanism matters more than the message. Explosive Media’s team confirmed to WIRED that content was prepared in advance for multiple political scenarios — Trump backing down, Trump escalating — requiring only minor adjustments before publication. Minutes after Trump’s Tuesday announcement that he would not destroy “a whole civilization,” a final-cut Lego video was live: Trump mini-figure seated beside a ceasefire document, holding a white flag, eating a taco. The TACO acronym — “Trump Always Chickens Out” — is an existing internet meme the team deliberately embedded, demonstrating the kind of cultural fluency that state media typically lacks. Iranian embassy accounts in Zimbabwe and Tunisia posted their own AI videos the same week, but with noticeably less precision.
Moustafa Ayad, a researcher with the Institute for Strategic Dialogue who has tracked Iranian online content throughout the conflict, told WIRED the videos work on two fronts simultaneously: they make the conflict legible from Iran’s perspective, and they map onto existing points of domestic disaffection in the United States. That dual targeting is not accidental — the team states that Americans themselves have been feeding them tips and ideas, which, if true, means the influence operation has achieved the hardest thing: organic audience co-production.
The governance question Ayad raises is blunt and evidence-grounded: Iran has effectively cut off internet access for its general population. Explosive Media claims it obtained access by registering as a media organisation with over 2.5 million followers across Iranian messaging channels. Ayad’s assessment — that operating at this level requires proximity to the government — has not been refuted by the group, only denied. The Iranian government has previously distributed Lego-style propaganda through Revolutionary Guard channels as far back as 2024. The aesthetic lineage is not ambiguous.
Atelier: Architecture and BIM practices are building AI visual pipelines right now — rendering, presentation, competition submissions. Explosive Media’s workflow is a compressed proof-of-concept: AI scripting plus AI animation plus rapid iteration equals broadcast-ready output in hours. The same pipeline that produces Lego Trump crying over a ceasefire document can produce a photorealistic residential scheme that never existed. The trust deficit this creates is a professional liability, not a political abstraction.
The IP angle deserves a line: Lego is a Danish company with active trademark protection over its minifigure aesthetic. No platform moderation response — from X, YouTube, or Telegram — has been reported regarding the Explosive Media content. The EU’s Digital Services Act obliges very large platforms to assess systemic risks from coordinated inauthentic behaviour; whether this campaign has triggered any DSA risk-assessment process is an open question that European regulators have not publicly answered.
For AEC professionals, the actionable read is this: verify the provenance of any AI-generated visual before it enters a client deliverable or a public competition submission. If a small anonymous team can produce millions-of-views content in hours with undisclosed AI tools, the visual credibility of synthetic imagery — in any domain — is now permanently contested. Build verification steps into your BIM Execution Plan and your media review workflow before a client or a Wettbewerb jury does it for you.
Source: Ars Technica
SOURCE · ↗