Residents of Tel Aviv are experiencing an unsettling phenomenon: watching AI-generated videos of their city in ruins while sitting in cafes overlooking intact skylines, highlighting how synthetic media is creating divergent realities of the Israel-Iran conflict.
"I'm sitting here in Tel Aviv, drinking a mediocre overpriced coffee, looking at a perfectly intact skyline," wrote one resident in a viral Reddit post. "Meanwhile, I hop on Reddit/X/TikTok and according to a bunch of 'OSINT' accounts (with anime profile pictures) and AI-generated videos, I'm currently typing this from a smoldering crater."
The disconnect between ground reality and social media narratives has reached unprecedented levels as artificial intelligence tools enable the rapid creation of convincing fake footage. Videos purporting to show Iranian missiles destroying Tel Aviv landmarks circulate widely despite live webcams showing normal urban activity, creating what digital media researchers describe as the first major conflict experiencing systematic AI-generated disinformation at scale.
The phenomenon extends beyond simple propaganda to reshape international perception of the conflict's intensity and impact. Accounts claiming "OSINT" (open source intelligence) credentials share AI-generated content alongside authentic footage, exploiting the difficulty non-experts face in distinguishing synthetic media from real documentation.
"You can literally go to YouTube, search for a 24/7 live cam of Tel Aviv, and see a bus stuck in traffic and a guy walking his Labradoodle," the resident noted. "But no, 'User_1234' from 5,000 miles away insists that Iran has 'leveled the Zionist entity' with a video where the missiles don't even have shadows."
Digital forensics experts point to telltale signs in AI-generated war footage: inconsistent lighting, physically impossible trajectories, and artifacts typical of generative models. Yet these technical indicators often go unnoticed by casual social media users, allowing fake content to achieve viral spread before fact-checkers can respond.

