By Mark Ginsberg
When the Missiles Stopped
When the missiles stopped falling in June 2025, Israel realized its toughest battle hadn’t been waged in the skies.
A deepfake depicting an F-35 being shot down circulated within hours of Iran’s actual strike. It wasn’t hard to spot if you knew what to look for, but most viewers didn’t. By the time Israel’s military confirmed the aircraft was safe, the clip had already been shared millions of times on Telegram, X, and TikTok. No correction could keep pace. The lie outran the truth, and the defenses designed to catch it were built for a different kind of war.
This was new.
The Detection Gap Nobody Planned For
Israel entered the June 2025 conflict with one of the strongest air defense systems in the world. Arrow, David’s Sling, and Iron Dome intercepted ballistic missiles and drones at rates analysts at CSIS called historic. What Israel lacked was a comparable defense for information itself. The country’s intelligence agencies were geared to track launch sites, troop movements, and weapons shipments, not to identify a synthetic video of an attack on the Dimona nuclear facility before it reached 2 million views.
The Israeli Internet Association documented over 400 distinct disinformation events during the 12-day conflict. These ranged from fabricated casualty numbers to manipulated satellite imagery showing nonexistent destruction in Tel Aviv. Their study found that Hebrew-language detection tools effectively did not exist. Facebook’s third-party fact-checking partners lacked sufficient Hebrew capacity, and Google’s AI content labels, rolled out months earlier for English and Portuguese markets, had not been extended to Hebrew or Arabic content from the region.
Israel was fighting an information war in languages its allies’ tools couldn’t understand. The same gap that FactSignal has been documenting in its coverage of Iranian disinformation campaigns, where state actors exploit platforms with the weakest detection infrastructure.
What the IDF Actually Changed
In August 2025, the IDF restructured its C4i Directorate.
The restructuring merged three previously separate units. Spectrum operations handled electronic warfare and signals. Lotem managed information technology. The strategic communications branch ran messaging. Each had historically operated in silos. For the first time, the military brought electronic warfare, cybersecurity, and information operations under a single command.
The reasoning was clear. During the war, fabricated content could originate from coordinated networks of accounts, travel through a compromised server overseas, and land on an Israeli citizen’s phone within minutes. Separate IDF units were responsible for different pieces of that puzzle, but none could connect them quickly enough to make a difference.
The restructuring addressed an organizational problem, not a technological one.
Israel possesses formidable signal intelligence and cybersecurity capabilities. What it lacked was a doctrine that treated information as a domain of warfare equal to air, sea, or land. The June 2025 conflict proved that a fabricated video could incite the same civilian panic as an air raid siren, yet the military had no equivalent of Iron Dome for content.
The Cost of Going It Alone
Israel’s experience isn’t unique in its essence. What sets it apart is the timing. The June 2025 war was the first major conflict to see generative AI tools readily available to both state actors and loosely organized influence networks from the outset. Iran didn’t require an elaborate propaganda machine. It needed a laptop, internet access, and open-source image generation models accessible to anyone.
The cost of producing convincing fabricated content dropped to near zero.
The cost of detecting it remained high and centralized. NATO’s StratCom Centre of Excellence estimated that verifying a single piece of AI-generated content during an active conflict takes an average of 72 hours for a trained analyst. Producing that same content takes minutes. This asymmetry represents the core challenge. No country has solved it.
Israel’s wartime adaptations, from the C4i merger to investments in Hebrew-language detection to partnerships with the Technion and Reichman University on automated content verification, all came after significant damage was done. The monitoring gap was identified in hindsight, the way most security failures are. PaxPoint has argued in its analysis of regional security trends that this adaptation gap is now a structural vulnerability for any allied nation operating in a contested information environment.
What Other Countries Should Learn
Countries observing Israel’s experience fall into two categories.
The first views the C4i restructuring as a model to study. Several Eastern European nations bordering Russia have launched similar internal reviews of their own information defense capabilities. Estonia, Latvia, and Poland each ran assessments in late 2025 examining whether their existing institutions could detect and respond to AI-generated content during an active military engagement.
The second camp considers Israel’s experience an outlier. These are countries that have not faced a recent conflict where information operations played a central role. They assume their existing media literacy programs and content moderation by tech companies will be sufficient. Israel shared that assumption until June 2025.
The key difference between Israel and most other democracies is that Israel got tested. The war forced a reckoning many governments haven’t faced.
This monitoring gap is not an Israeli problem. It is a timing issue. Every country will eventually encounter a conflict where AI-generated content is part of the weapons mix. The question is whether they invest in detection capabilities before or after the first fabricated video reaches their citizens. Allyvia has covered this dynamic extensively, particularly how democracies that assume their institutions will hold often discover the cracks only under pressure.
Israel built it after. The cost of that delay was measurable in panicked evacuations, diplomatic fallout from fabricated statements attributed to Israeli officials, and a temporary but real erosion of public trust in online information during a war when accurate information meant life and death.
No country should want to learn that lesson the hard way.
Author: Mark Ginsberg – Israel-based writer and commentator on technology and artificial intelligence, with over two decades of firsthand perspective on the country’s innovation economy. Originally from South Florida and now based in Efrat, he analyzes the growth of Israel’s tech ecosystem and the broader implications of AI on business, policy, and society. His work has appeared in Small Wars Journal and Breaking Defense.
(The opinions expressed in this article belong only to the author and do not necessarily reflect the views of World Geostrategic Insights).
Image Source: IDF (Officers in the new “Alumot” unit in the C4I and Cyber Defense Directorate).