Donald Trump Is the First AI Slop President (www.wired.com)

President Donald Trump and his team have been circulating a string of crude, AI-generated videos on his main accounts — from a Top Gun–style clip of him dumping excrement on fictional protesters to racist caricatures of political opponents — sometimes posted by staff after he saves or approves them. Sources say Trump often isn’t creating the clips himself; two aides, Dan Scavino and Natalie Harp, are typically responsible for executing posts on Truth Social, though the administration is evasive about who produced or authorized specific videos. The content is lowbrow and inflammatory rather than the sophisticated “deepfake” many feared, but it’s clearly engineered with generative video tools and distributed from high-profile channels. For the AI/ML community this is a wake-up call about real-world misuse: synthetic video is being normalized as political trolling, lowering the bar for harmful imagery and complicating provenance, moderation, and attribution. Key technical unknowns remain — which models or services (e.g., xAI’s Grok Imagine, OpenAI’s Sora, or other tools) generated the material, and what workflows were used to vet and post it. Even if not state-of-the-art deepfakes, these clips highlight urgent needs for robust provenance metadata, reliable detection methods, platform policy enforcement, and clearer responsibility chains to prevent erosion of trust and guardrails against more consequential manipulations.