The European Commission published its final Guidelines on transparency obligations under Article 50 of the EU AI Act on 20 July 2026.
Although the Guidelines are non-binding, they offer significant practical clarification for businesses ahead of the formal application of these obligations.
Article 50 is not limited to high-risk AI systems and may apply across a broad range of AI solutions used by businesses today.
Affected technologies include chatbots, generative AI tools, emotion recognition systems, biometric categorisation tools, and deepfake technologies.
The core purpose of Article 50 is to reduce transparency risks such as impersonation, deception, manipulation, and misinformation when individuals interact with AI systems.
The Guidelines draw a clear distinction between providers, who develop or place AI systems on the market, and deployers, who use AI systems under their own authority.
Four main transparency scenarios are identified, each placing specific disclosure duties on either providers or deployers depending on the context of use.
Providers of interactive AI systems must inform individuals when they are interacting with AI, unless this is already obvious from the surrounding context.
Providers of systems that generate synthetic audio, image, video, or text content must ensure outputs are marked in a machine-readable format and detectable as artificially generated or manipulated.
Deployers using emotion recognition or biometric categorisation systems must inform exposed individuals, subject to applicable prohibitions and data protection requirements.
Deployers must also disclose deepfake content and, in certain cases, AI-generated text published to inform the public on matters of public interest.
Transparency information must be clear, distinguishable, and provided no later than the first interaction or exposure to AI-generated content.
Generic references hidden in terms and conditions, website footers, or vague labels are unlikely to be considered sufficient under the Guidelines.
The territorial scope of the obligations is broad and may extend to non-EU providers and deployers where AI outputs are used within the EU.
However, incidental, unforeseeable, or unauthorised downstream use should not alone trigger obligations for third-country providers operating outside the bloc.
Businesses are now advised to map their AI use cases and determine whether they act as a provider, deployer, or both under the framework.
Organisations should also review user-facing disclosures, assess marking requirements for AI-generated content, and document any reliance on available exemptions.
Law firm K&L Gates, whose lawyers Sarah Pearce and Veronica Muratori authored analysis of the Guidelines, highlighted the wide-ranging implications for businesses operating AI systems across multiple sectors.
Companies with EU exposure should treat these Guidelines as an urgent prompt to audit current AI deployment practices before obligations take full effect.