DAVID-XR1: Detecting AI-Generated Videos with Explainable Reasoning

TL;DR

This paper introduces DAVID-XR1, a novel video-language model that provides explainable, fine-grained detection of AI-generated videos using detailed annotations and reasoning, enhancing trustworthiness and transparency.

Contribution

It presents DAVID-XR1, the first model to offer interpretable, defect-level explanations for AI-generated video detection, supported by a new dataset with rich annotations.

Findings

Model achieves strong generalization across different AI video generators.

Fine-grained explanations improve trust and interpretability.

Chain-of-thought distillation enhances detection accuracy.

Abstract

As AI-generated video becomes increasingly pervasive across media platforms, the ability to reliably distinguish synthetic content from authentic footage has become both urgent and essential. Existing approaches have primarily treated this challenge as a binary classification task, offering limited insight into where or why a model identifies a video as AI-generated. However, the core challenge extends beyond simply detecting subtle artifacts; it requires providing fine-grained, persuasive evidence that can convince auditors and end-users alike. To address this critical gap, we introduce DAVID-X, the first dataset to pair AI-generated videos with detailed defect-level, temporal-spatial annotations and written rationales. Leveraging these rich annotations, we present DAVID-XR1, a video-language model designed to deliver an interpretable chain of visual reasoning-including defect categorization, temporal-spatial localization, and natural language explanations. This approach fundamentally transforms AI-generated video detection from an opaque black-box decision into a transparent and verifiable diagnostic process. We demonstrate that a general-purpose backbone, fine-tuned on our compact dataset and enhanced with chain-of-thought distillation, achieves strong generalization across a variety of generators and generation modes. Our results highlight the promise of explainable detection methods for trustworthy identification of AI-generated video content.