The EU AI Act AI content marking rule has a new deadline that matters to many businesses: December 2, 2026. Regulation (EU) 2026/1744, known as the Digital Omnibus on AI, was published in the Official Journal of the European Union on July 24, 2026 and entered into force on July 27. It gives providers of generative AI systems already on the market before August 2, 2026 until December 2, 2026 to meet the Article 50(2) duty to mark AI-generated content, and it pushes back several high-risk deadlines.
This article is general information, not legal advice.
What changed
Article 50(2) of the AI Act requires providers of AI systems that generate synthetic audio, images, video or text to embed machine-readable markers in that content, so it can be detected as AI-generated. The general application date for Article 50 is August 2, 2026. Under the Omnibus, generative AI systems placed on the market before that date get a transitional period that runs until December 2, 2026. Law firm Gibson Dunn describes this as a four-month grace period. Systems placed on the market after August 2, 2026 do not get that extra time.
The other Article 50 transparency duties, such as telling people when they are interacting with an AI system, were not deferred and apply from August 2, 2026. High-risk rules moved further out. Obligations for stand-alone high-risk systems listed in Annex III, which covers areas such as biometrics, employment, education and access to essential services, now apply from December 2, 2027 instead of August 2, 2026. Rules for high-risk AI embedded in regulated products under Annex I now apply from August 2, 2028.
The Omnibus also softens the AI literacy duty in Article 4, so providers and deployers must take measures to support staff AI literacy without a fixed required level. It extends some benefits previously reserved for SMEs, such as simplified technical documentation, to small mid-cap companies, and adds new prohibitions, including on AI systems that generate non-consensual intimate images of identifiable people.
Why the EU AI Act AI content marking deadline matters for businesses
The marking duty sits with providers, the companies that develop generative AI systems and put them on the market. Most small and mid-size businesses that use AI to write blog posts, generate product images or draft emails are deployers, not providers, so the technical marking work falls mainly on their software vendors. That does not make the deadline irrelevant. If your vendor’s tool cannot mark its output, you inherit the operational problem.
The picture changes if you build AI into your own product. A software company, ecommerce brand or agency that offers customers a feature that generates text, images or audio may count as a provider of that system. Businesses outside the EU, including those in the US, GCC and India, that offer AI features to customers in the EU should check their position as well.
The high-risk delay is real relief for companies using AI in hiring, credit or education decisions, but the extra time is best used for preparation, not postponement. The literacy and transparency duties that already apply still need attention now.
What to do now
- Map every AI tool your business uses to generate content, and note whether you are using a vendor’s system or offering your own AI feature to customers.
- Ask each generative AI vendor in writing how and when its output will carry machine-readable marking under Article 50(2).
- If you provide a generative AI feature in the EU, confirm when it was placed on the market, since that determines whether the December 2, 2026 transitional date applies.
- Review chatbots and other customer-facing AI for clear disclosure that people are interacting with AI, a duty that already applies.
- Keep a simple AI literacy record: who uses which tools, what training they received and which internal guidelines apply.
- Get qualified legal advice on your specific role and obligations before the deadline.
Purple Rocket works with businesses across the EU on websites and content, and our AI integration services team can help you build AI features with transparency in mind.