If your business uses AI in marketing, customer service or content creation, the EU AI Act is already relevant to you. Its Article 50 transparency rules have applied since 2 August 2026, reaching beyond technology companies to businesses and individuals using AI professionally in the EU.
The obligations depend on your role. Businesses using third-party tools are typically ‘deployers’; those developing or commissioning systems under their own name may also be ‘providers’, with additional responsibilities. Purely personal, non-professional use is excluded.
When must AI use be disclosed?
Providers must generally ensure people are told when they interact directly with AI, unless this is obvious. Providers of generative AI systems also face technical duties to mark outputs in machine-readable form. These are distinct from the visible or audible disclosures that businesses may need to make when using AI-generated content.
For marketing teams, the key question is whether images, audio or video qualify as deepfakes. Realistic synthetic people, voices or product presentations may fall within the rules if they falsely appear authentic. Where disclosure is required, it must be clear by the audience’s first exposure; hidden metadata alone is insufficient.
Not every AI-assisted text needs a label. The text-disclosure rule concerns publication to inform the public on matters of public interest. An exemption applies where there is substantive human review or editorial control and a person or organisation takes editorial responsibility. A spelling check is not enough.[1]
Check the timetable
July 2026 amendments postponed the rules for high-risk uses such as recruitment to 2 December 2027, and certain product-related high-risk rules to 2 August 2028.[2]
This is not a general postponement of transparency duties. The grace period until 2 December 2026 applies only to providers’ marking and detection duties for systems placed on the market before 2 August 2026. It does not apply to businesses’ deepfake disclosures.
Protect the value of your content
An AI label is not permission to use another party’s protected material, brand or likeness. Nor does it establish copyright in the output: human creative contribution remains essential. Businesses should check rights and supplier terms, and document that contribution before treating AI-generated content as an exclusive asset.[3]
We recommend mapping AI use, assigning responsibility, reviewing agency contracts and introducing publication checks and staff training. For owners and investors, these are also due-diligence questions about reputation and asset value. Transparency breaches can attract fines of up to €15 million or 3% of worldwide annual turnover, with specific protections for smaller businesses.
Should you wish to know more about how the AI Act affects your business, or require assistance with compliance, contracts or protecting your intellectual property, please do not hesitate to contact us.
[1] European Commission, Transparency obligations under Article 50 of the AI Act (FAQs, 24 July 2026).
[2] European Commission, AI Omnibus enters into force (27 July 2026).
[3] EUIPO, Generative AI in Education: Understanding copyright implications.
