EU Unveils Voluntary Code to Enhance Transparency in AI-Generated Content
Nadia Petrov | AA
The European Commission has published a Code of Practice to label AI-generated content, including deepfakes, ahead of mandatory transparency measures under the EU AI Act starting August 2.
The European Commission has published the Code of Practice on marking and labeling artificial intelligence (AI)-generated content, a significant voluntary framework designed to support compliance with the EU AI Act’s impending transparency requirements. These measures, which will become legally enforceable on August 2, reflect the European Union’s commitment to mitigating risks associated with misinformation and deceptive practices in an increasingly AI-dominated digital space.
The Code of Practice aims to ensure that both providers and deployers label AI-generated or manipulated content such as text, audio, images, and videos clearly and in a machine-readable format. According to the transparency obligations, deepfakes must be identified, and users must be informed whenever they are interacting with an AI system, including chatbots. The Commission emphasized that these provisions are crucial tools for empowering citizens to understand the origins of information and discern whether it has been artificially created or manipulated.
How the Code Targets Providers and Deployers of AI Systems
The Code of Practice features separate directives for two distinct stakeholder groups: providers and deployers of AI systems. For providers, guidelines include marking AI-generated or AI-altered data, ensuring it can be detected as artificial, and adhering to machine-readable formatting standards. Deployers, on the other hand, are tasked with labeling deepfakes and AI-manipulated content published to inform the public, particularly in scenarios where no human review or editorial oversight occurs. By doing so, the EU Commission aims to close gaps in accountability and prevent misleading information from spreading unchecked.
The transparency measures represent only one aspect of the broader EU AI Act. The legislation also addresses general-purpose AI systems and those deemed high-risk to support responsible technology development across the bloc. In welcoming providers and deployers to join the initiative as signatories, EU officials underline the collaborative intention behind the guidelines, which aim to boost industry alignment with ethical AI practices.
Next Steps for Implementation and Compliance
The EU Commission revealed that signatories to the Code of Practice will demonstrate their compliance with AI Act requirements once the initiative is approved by the Commission and the AI Board. To complement the code, accompanying guidelines will be issued to clarify the scope of the mandated legal provisions, as well as to address areas not explicitly covered in the document. These steps are part of a broader effort to ensure seamless integration of transparency measures across various sectors where artificial intelligence plays a growing role.
Transparency Rules Signal a Shift in AI Regulation for Cities
The publication of the Code of Practice represents a pivotal moment in how cities and communities within the European Union handle artificial intelligence’s impact on public-interest communications. As urban centers increasingly experiment with AI-driven systems, including chatbots for municipal services, ensuring trust and accountability becomes critical to maintaining civic engagement. The labeling requirements for AI-generated content aim to fortify the lines between human and machine interactions, protecting individuals from unknowingly being subjected to manipulated information or synthetic media.
City leaders should view these developments as an opportunity to proactively align local policies with the EU’s broader goals of transparency and ethical AI governance. By incorporating clear labeling systems into municipal operations,be it in public service announcements or automated administrative details,cities can mitigate reputational risks and strengthen their relationship with residents. Transparency, as underscored by the Code of Practice, is not just a regulatory necessity but an essential foundation for public trust in the ever-evolving digital age.
Questions That Matter
What is the EU’s Code of Practice for AI labeling?
The Code of Practice outlines voluntary measures for marking AI-generated content to comply with EU AI Act transparency obligations. It applies to providers and deployers ahead of mandatory rules starting August 2.
What does the AI Act require by August 2?
The AI Act mandates clear labeling of AI-generated or manipulated content and alerts for public engagement with AI systems, focusing on transparency to reduce risks of deception and misinformation.