EU Unveils AI Labeling Playbook Ahead of August Deadline
The European Union has launched a voluntary Code of Practice to guide companies on labeling AI-generated content, aiming to align with the AI Act’s transparency rules set to take effect on August 2, 2026. Released on June 10, the Code offers a framework for businesses to mark AI outputs, though compliance with its mandates is mandatory under Article 50 of the AI Act. While signing the Code is optional, the obligations it outlines—such as labeling deepfakes and AI-altered text—must be followed regardless of participation.
Mandatory Transparency Rules Under the AI Act
From August 2, 2026, two key requirements will apply: AI-generated content on public-interest topics must carry a visible label, and users interacting with AI systems (like chatbots) must be informed they are engaging with a machine. The European Commission emphasizes these measures to combat deception and empower users to discern AI’s role in shaping public discourse. “Europeans have a right to know whether content has been made or altered by AI,” said Henna Virkkunen, the Commission’s executive vice-president.
Collaborative Approach to Labeling
The Code splits responsibilities between AI developers and deployers. Model builders must embed machine-readable labels for AI outputs, while companies using AI in products handle visible labels. A unified EU icon is proposed to ensure consistency, reducing the burden on businesses to create their own. The Commission also encourages adoption of open technical standards to streamline implementation.
Rushed Timeline and Pending Guidelines
With less than two months until the August deadline, companies must swiftly adapt to labeling requirements. While the Code provides a starting point, further clarity will come from upcoming Commission guidelines. The Code, developed by six experts with input from 180 stakeholders, remains open for signatures and awaits final approval from the AI Board. Businesses now face a tight window to prepare, as the EU tightens its approach to AI accountability.
Source: AI News. AI-assisted editorial synthesis — TechnoExpress.