A Multi-Level Strategy for Deepfake Content Moderation under EU Regulation

2025-07-10

AI Summary

Addressing the lack of standardized frameworks, methodological rigidity, and poor scalability in deepfake governance under EU AI regulation, this paper proposes a multi-tiered collaborative governance framework. The framework integrates content labeling, detection, and annotation techniques, and introduces two key innovations: (i) a contextualized risk-weighting mechanism that dynamically assesses threat severity based on deployment context, and (ii) a lightweight, model-agnostic scoring system enabling cross-platform interoperability without reliance on specific generative models. Grounded in a systematic review of multi-source regulatory literature—including EU transparency requirements and AI Act compliance mandates—the framework establishes a compliant, scalable, and deployable content assessment architecture. Experimental evaluation demonstrates substantial improvements in feasibility for large-scale online content moderation and robust cross-platform adaptability, effectively bridging critical gaps in regulatory alignment and operational practicality of existing approaches.

Technology Category

• Philosophy and Ethics of AI: AI & Law, Justice, Regulation & Governance
• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Generative Adversarial Networks (GANs) for Vision

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The growing availability and use of deepfake technologies increases risks for democratic societies, e.g., for political communication on online platforms. The EU has responded with transparency obligations for providers and deployers of Artificial Intelligence (AI) systems and online platforms. This includes marking deepfakes during generation and labeling deepfakes when they are shared. However, the lack of industry and enforcement standards poses an ongoing challenge. Through a multivocal literature review, we summarize methods for marking, detecting, and labeling deepfakes and assess their effectiveness under EU regulation. Our results indicate that individual methods fail to meet regulatory and practical requirements. Therefore, we propose a multi-level strategy combining the strengths of existing methods. To account for the masses of content on online platforms, our multi-level strategy provides scalability and practicality via a simple scoring mechanism. At the same time, it is agnostic to types of deepfake technology and allows for context-specific risk weighting.

Problem

• Addressing risks of deepfake technologies in democratic societies
• Developing standards for deepfake marking and labeling under EU regulation
• Proposing a scalable multi-level strategy for deepfake content moderation

Innovation

• Multi-level strategy combining existing methods
• Scalable scoring mechanism for content moderation
• Technology-agnostic with context-specific risk weighting