UK AI Security Institute Brief

Headline

European Commission publishes voluntary Code of Practice ahead of AI Act transparency obligations effective 2 August 2026

Executive Summary

On 10 June 2026, the European Commission published the final Code of Practice on marking and labelling of AI-generated content. The Code offers voluntary implementation guidance for providers and deployers of generative AI systems before the binding transparency requirements of the AI Act take effect on 2 August 2026.

Key Regulatory Signals

• Binding Deadline: AI Act transparency obligations apply from 2 August 2026. Deepfakes, AI-generated or AI-manipulated text on matters of public interest, and interactive AI systems such as chatbots must carry clear labelling.
• Voluntary Status: The Code is non-binding. It sets out practical steps providers and deployers may adopt to demonstrate alignment with the AI Act's Article 50 transparency requirements before the August deadline.
• EU Standardised Icons: The Commission has published a dedicated EU icon set for labelling AI-generated content, available at digital-strategy.ec.europa.eu, providing a common visual standard signatories may adopt.
• Signature Process: A formal sign-on mechanism and an information session on the signature process have been published, enabling providers and deployers to register adherence to the Code ahead of the effective date.

Regulatory Delta

- The Code sits within the AI Act's phased application schedule. Article 50 transparency obligations represent the first wave of binding requirements to take effect, following the general-purpose AI model rules applicable from August 2025 and the prohibited-practices provisions that applied from February 2026.

- The voluntary Code is structurally new. No equivalent EU-level labelling code for AI-generated content existed before this publication, making it a first-in-kind instrument under the AI Act framework.

- The Digital Services Act (Regulation (EU) 2022/2065) already imposes transparency obligations on very large online platforms regarding AI-generated content in advertising. Article 50 of the AI Act extends labelling duties to a broader population of providers and deployers beyond those designated under the DSA.

Materiality Classification

HIGH — A binding regulatory deadline of 2 August 2026 requires providers and deployers of generative AI systems operating in the EU to have compliant labelling and disclosure mechanisms in place; the obligation applies sector-wide across all covered entities, not solely to named respondents.

Time Horizon

effective — 2026-08-02

Intelligence Outlook

Monitor the European Commission's digital strategy portal and the AI Office for the list of Code signatories, any clarificatory guidance on Article 50 scope, and enforcement posture signals from national market surveillance authorities after 2 August 2026.