Synthetic War: X Moderation Algorithms Surrender to Deepfakes

On March 15, 2026, the critical inability of social networks to counter machine-generated propaganda was exposed. The X platform (formerly Twitter) recorded a massive spread of fake videos realistically depicting nonexistent combat operations in the Middle East.

Despite the introduction of new corporate policies for labeling synthetic content, the system is not working. This is a purely technical problem: the quality of generation (from Midjourney V7 and Sora) has reached a level where the platform's algorithmic detectors produce a huge number of false negatives, and Community Notes lag behind due to viral spreading speeds. This precedent proves that in the information war of 2026, the attacker (the deepfake generator) has an asymmetric technological advantage over the defender (the platform moderator).

Source: Daily Sabah