The specific consequence of a footage-first conversation is a public that has calibrated its expectations around what AI-assisted strikes look like — precision, speed, minimal visible chaos — without any exposure to the decision architecture behind them. Strike videos from Marwahin [1] and Kafr Rumman [2] circulate in communities that are fluent in military hardware and operational context, but that fluency is visual and tactical, not legal or systemic. The audiences can identify the munition. They cannot see the targeting pipeline that selected the building. That asymmetry is not a…