[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$f4IBu-K7A96sdKZDs6g1WZ6g-1XKm9Nf22TClyvMdsuw":3,"faq-section-":72},{"tableOfContents":4,"markDownContent":5,"htmlContent":6,"metaTitle":7,"metaDescription":8,"wordCount":9,"readTime":10,"title":11,"nbDownloads":12,"excerpt":13,"lang":14,"url":15,"intro":16,"featured":4,"state":17,"author":18,"authorId":19,"datePublication":23,"dateCreation":24,"dateUpdate":25,"mainCategory":26,"categories":33,"metaDatas":36,"imageUrl":62,"imageThumbUrls":63,"id":71},false,"On 2 December 2026, the transitional period under Article 50(2) of the AI Act ends, and two new prohibitions under Article 5 become applicable. These deadlines do not affect all organisations in the same way, nor do they mark the start of all obligations under the Regulation.\n\nThis article explains what is changing and offers a **free downloadable checklist** to help you prepare.\n\n## What happens on 2 December 2026 under the AI Act?\n\nThree requirements apply from 2 December 2026, pursuant to Regulation (EU) 2024/1689 as amended by [Regulation (EU) 2026/1744](https://eur-lex.europa.eu/eli/reg/2026/1744/oj/eng), known as the “Digital Omnibus on AI”.\n\n| What becomes applicable | Who is concerned | Legal basis |\n| --- | --- | --- |\n| Machine-readable marking of synthetic content (text, images, audio and video) | Providers of systems placed on the market before 2 August 2026 | Article 50(2) and Article 111(4) |\n| Prohibition of AI systems that generate or manipulate realistic intimate depictions of an identifiable person without that person’s consent | Providers, and deployers using a system for that purpose | Article 5(1)(b) |\n| Prohibition of AI systems that generate or manipulate child sexual abuse material | Providers, and deployers using a system for that purpose | Article 5(1)(b) |\n\n## Article 50(2): the obligation to mark AI-generated content\n\nArticle 50(2) requires providers of generative AI systems to mark their outputs in a machine-readable format so that they can be detected as artificially generated or manipulated. This is not a new general obligation for every organisation using AI: it imposes specific obligations on providers.\n\n2 December 2026 is the deadline for systems placed on the market **before** 2 August 2026 (Article 111(4)). Systems launched from 2 August 2026 onwards do not benefit from any transitional period.\n\n**What the Regulation requires:**\n\n- effective, interoperable, robust and reliable marking solutions, insofar as technically feasible;\n- techniques such as signed provenance metadata or imperceptible watermarks, often used in combination;\n- a means of detecting that content originates from the system;\n- compliance with the exemptions and technical conditions provided for by the Regulation, for example for standard editing assistance functions.\n\nRobust text marking remains an open technical issue: the feasibility assessment should be documented. Non-compliance with Article 50 may result in fines of up to EUR 15 million or 3% of worldwide annual turnover.\n\n## The two new prohibited practices under Article 5\n\nArticle 5 now also prohibits AI systems that generate or manipulate realistic intimate depictions of an identifiable person without their explicit consent, as well as systems that generate child sexual abuse material. An infringement of Article 5 remains subject to fines of up to EUR 35 million or 7% of worldwide annual turnover.\n\nThe scope depends on your role:\n\n- **Provider:** the prohibition applies where such generation is the intended purpose of the system, or a reasonably foreseeable and reproducible outcome where technical safeguards are insufficient. It may therefore apply to general-purpose generative AI systems.\n- **Deployer:** the prohibition applies where the deployer uses a system to generate or manipulate such content.\n\nExpected safeguards range from training data cleansing to output filters, refusal training, abuse detection, and reporting and takedown mechanisms. The exact scope, required consent and exemptions are set out in Article 5.\n\n## What does not change on 2 December 2026\n\n2 December 2026 is not the date on which all AI Act obligations begin to apply, and it is best not to present it as such internally.\n\n- The transparency obligations under Article 50 have applied since 2 August 2026. Only the marking requirement under Article 50(2) benefits from a transitional period, for systems already placed on the market on that date.\n- Obligations relating to high-risk AI systems apply from 2 December 2027 for Annex III systems, and from 2 August 2028 for Annex I systems.\n- The remainder of Article 5 has applied since 2 February 2025.\n\n## What the downloadable checklist includes\n\nThe checklist provides practical actions in the form of tick-box steps, together with the specific checks to be carried out for each one.\n\n**You will find:**\n\n- six steps with tick boxes: role, scope, timeline, technical controls, prohibited uses and evidence;\n- an evidence-retention table for each system;\n- a three-week action plan leading up to 2 December;\n- useful legal references.\n\n**Who it is for:** Privacy and DPO teams, AI Compliance leads, legal teams and product teams, whether they are providers or deployers of AI systems.\n\n## Are you ready for 2 December 2026?\n\nDownload the free checklist to assess your position system by system and identify what needs to be documented before the deadline.\n\n### Prepare now with Dastra\n\nDastra helps Privacy and AI Compliance teams manage their AI Act compliance on a single platform: inventory your AI systems, identify roles and scope, assess your providers, track risks and measures, build your evidence file, and link it all to your GDPR compliance.\n\n[Book a demo here.](https://www.dastra.eu/en/contacts/demo)","\u003Cp>On 2 December 2026, the transitional period under Article 50(2) of the AI Act ends, and two new prohibitions under Article 5 become applicable. These deadlines do not affect all organisations in the same way, nor do they mark the start of all obligations under the Regulation.\u003C/p>\n\u003Cp>This article explains what is changing and offers a \u003Cstrong>free downloadable checklist\u003C/strong> to help you prepare.\u003C/p>\n\u003Ch2 id=\"what-happens-on-2-december-2026-under-the-ai-act\">What happens on 2 December 2026 under the AI Act?\u003C/h2>\n\u003Cp>Three requirements apply from 2 December 2026, pursuant to Regulation (EU) 2024/1689 as amended by \u003Ca href=\"https://eur-lex.europa.eu/eli/reg/2026/1744/oj/eng\" rel=\"nofollow\">Regulation (EU) 2026/1744\u003C/a>, known as the “Digital Omnibus on AI”.\u003C/p>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>What becomes applicable\u003C/th>\n\u003Cth>Who is concerned\u003C/th>\n\u003Cth>Legal basis\u003C/th>\n\u003C/tr>\n\u003C/thead>\n\u003Ctbody>\n\u003Ctr>\n\u003Ctd>Machine-readable marking of synthetic content (text, images, audio and video)\u003C/td>\n\u003Ctd>Providers of systems placed on the market before 2 August 2026\u003C/td>\n\u003Ctd>Article 50(2) and Article 111(4)\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>Prohibition of AI systems that generate or manipulate realistic intimate depictions of an identifiable person without that person’s consent\u003C/td>\n\u003Ctd>Providers, and deployers using a system for that purpose\u003C/td>\n\u003Ctd>Article 5(1)(b)\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>Prohibition of AI systems that generate or manipulate child sexual abuse material\u003C/td>\n\u003Ctd>Providers, and deployers using a system for that purpose\u003C/td>\n\u003Ctd>Article 5(1)(b)\u003C/td>\n\u003C/tr>\n\u003C/tbody>\n\u003C/table>\n\u003Ch2 id=\"article-502-the-obligation-to-mark-ai-generated-content\">Article 50(2): the obligation to mark AI-generated content\u003C/h2>\n\u003Cp>Article 50(2) requires providers of generative AI systems to mark their outputs in a machine-readable format so that they can be detected as artificially generated or manipulated. This is not a new general obligation for every organisation using AI: it imposes specific obligations on providers.\u003C/p>\n\u003Cp>2 December 2026 is the deadline for systems placed on the market \u003Cstrong>before\u003C/strong> 2 August 2026 (Article 111(4)). Systems launched from 2 August 2026 onwards do not benefit from any transitional period.\u003C/p>\n\u003Cp>\u003Cstrong>What the Regulation requires:\u003C/strong>\u003C/p>\n\u003Cul>\n\u003Cli>effective, interoperable, robust and reliable marking solutions, insofar as technically feasible;\u003C/li>\n\u003Cli>techniques such as signed provenance metadata or imperceptible watermarks, often used in combination;\u003C/li>\n\u003Cli>a means of detecting that content originates from the system;\u003C/li>\n\u003Cli>compliance with the exemptions and technical conditions provided for by the Regulation, for example for standard editing assistance functions.\u003C/li>\n\u003C/ul>\n\u003Cp>Robust text marking remains an open technical issue: the feasibility assessment should be documented. Non-compliance with Article 50 may result in fines of up to EUR 15 million or 3% of worldwide annual turnover.\u003C/p>\n\u003Ch2 id=\"the-two-new-prohibited-practices-under-article-5\">The two new prohibited practices under Article 5\u003C/h2>\n\u003Cp>Article 5 now also prohibits AI systems that generate or manipulate realistic intimate depictions of an identifiable person without their explicit consent, as well as systems that generate child sexual abuse material. An infringement of Article 5 remains subject to fines of up to EUR 35 million or 7% of worldwide annual turnover.\u003C/p>\n\u003Cp>The scope depends on your role:\u003C/p>\n\u003Cul>\n\u003Cli>\u003Cstrong>Provider:\u003C/strong> the prohibition applies where such generation is the intended purpose of the system, or a reasonably foreseeable and reproducible outcome where technical safeguards are insufficient. It may therefore apply to general-purpose generative AI systems.\u003C/li>\n\u003Cli>\u003Cstrong>Deployer:\u003C/strong> the prohibition applies where the deployer uses a system to generate or manipulate such content.\u003C/li>\n\u003C/ul>\n\u003Cp>Expected safeguards range from training data cleansing to output filters, refusal training, abuse detection, and reporting and takedown mechanisms. The exact scope, required consent and exemptions are set out in Article 5.\u003C/p>\n\u003Ch2 id=\"what-does-not-change-on-2-december-2026\">What does not change on 2 December 2026\u003C/h2>\n\u003Cp>2 December 2026 is not the date on which all AI Act obligations begin to apply, and it is best not to present it as such internally.\u003C/p>\n\u003Cul>\n\u003Cli>The transparency obligations under Article 50 have applied since 2 August 2026. Only the marking requirement under Article 50(2) benefits from a transitional period, for systems already placed on the market on that date.\u003C/li>\n\u003Cli>Obligations relating to high-risk AI systems apply from 2 December 2027 for Annex III systems, and from 2 August 2028 for Annex I systems.\u003C/li>\n\u003Cli>The remainder of Article 5 has applied since 2 February 2025.\u003C/li>\n\u003C/ul>\n\u003Ch2 id=\"what-the-downloadable-checklist-includes\">What the downloadable checklist includes\u003C/h2>\n\u003Cp>The checklist provides practical actions in the form of tick-box steps, together with the specific checks to be carried out for each one.\u003C/p>\n\u003Cp>\u003Cstrong>You will find:\u003C/strong>\u003C/p>\n\u003Cul>\n\u003Cli>six steps with tick boxes: role, scope, timeline, technical controls, prohibited uses and evidence;\u003C/li>\n\u003Cli>an evidence-retention table for each system;\u003C/li>\n\u003Cli>a three-week action plan leading up to 2 December;\u003C/li>\n\u003Cli>useful legal references.\u003C/li>\n\u003C/ul>\n\u003Cp>\u003Cstrong>Who it is for:\u003C/strong> Privacy and DPO teams, AI Compliance leads, legal teams and product teams, whether they are providers or deployers of AI systems.\u003C/p>\n\u003Ch2 id=\"are-you-ready-for-2-december-2026\">Are you ready for 2 December 2026?\u003C/h2>\n\u003Cp>Download the free checklist to assess your position system by system and identify what needs to be documented before the deadline.\u003C/p>\n\u003Ch3 id=\"prepare-now-with-dastra\">Prepare now with Dastra\u003C/h3>\n\u003Cp>Dastra helps Privacy and AI Compliance teams manage their AI Act compliance on a single platform: inventory your AI systems, identify roles and scope, assess your providers, track risks and measures, build your evidence file, and link it all to your GDPR compliance.\u003C/p>\n\u003Cp>\u003Ca href=\"https://www.dastra.eu/en/contacts/demo\">Book a demo here.\u003C/a>\u003C/p>\n","AI Act: Are You Ready for 2 December 2026?","Are you ready for 2 December 2026? Article 50(2), new prohibitions and 6 checks to carry out. Download the free AI Act checklist.",829,5,"AI Act: Is Your Organization Ready for December 2, 2026?",0,null,"en","ai-act-checklist-en","Are you ready for 2 December 2026? Article 50(2), new prohibitions, and 6 checks to carry out. Download the free AI Act checklist.","Published",{"id":19,"displayName":20,"avatarUrl":21,"bio":13,"blogUrl":13,"color":13,"userId":19,"creationDate":22},20352,"Leïla Sayssa","https://static.dastra.eu/tenant-3/avatar/20352/TDYeY3C8Rz1lLE/dpo-avatar-h01-150.png","2025-03-03T11:08:22","2026-11-17T15:07:00","2026-11-17T15:07:01.2688405","2026-11-19T12:41:21.9944427",{"id":27,"name":28,"description":13,"url":29,"color":30,"parentId":13,"count":13,"imageUrl":13,"parent":13,"order":31,"translations":32},70,"Livre blanc","white-papers","#1795d3",3,[],[34],{"id":27,"name":28,"description":13,"url":29,"color":30,"parentId":13,"count":13,"imageUrl":13,"parent":13,"order":31,"translations":35},[],[37,41],{"typeMetaDataId":38,"value":39,"id":40},4,"https://static.dastra.eu/backofficefilescontainer/bc880b52-7ec0-4e3d-88cc-ab153c4c0e40/Checklist AI Act 2 December - Dastra EN.pdf",120483,{"typeMetaDataId":42,"value":43,"id":61},9,{"title":44,"items":45},"Frequently asked questions about the AI Act and the Dec. 2, 2026 deadline",[46,49,52,55,58],{"question":47,"answer":48},"Which obligations apply on 2 December 2026?","\u003Cp>The machine-readable marking of synthetic content (Art. 50(2)) for systems placed on the market before 2 August 2026, and two new prohibitions under Article 5.\u003C/p>\u003Cp>\u003C/p>",{"question":50,"answer":51},"Is a deployer concerned by Article 50(2)?","\u003Cp>No, this obligation falls on providers. Deployers have other transparency obligations under Article 50, applicable since 2 August 2026, and remain subject to the Article 5 prohibitions.\u003C/p>\u003Cp>\u003C/p>",{"question":53,"answer":54},"My system was launched after 2 August 2026: do I get a grace period?","\u003Cp>No. The 2 December 2026 deadline only covers systems already on the market before 2 August 2026.\u003C/p>\u003Cp>\u003C/p>",{"question":56,"answer":57},"What are the penalties for non-compliance?","\u003Cp>An infringement of Article 5 can reach €35 million or 7% of worldwide turnover. For Article 50, the cap is €15 million or 3%.\u003C/p>\u003Cp>\u003C/p>",{"question":59,"answer":60},"Do the high-risk system obligations also apply?","\u003Cp>No. They apply from 2 December 2027 for Annex III and from 2 August 2028 for Annex I.\u003C/p>",120484,"https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance-original.png",[64,65,66,67,68,69,70],"https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance-1000.webp","https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance.webp","https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance-1500.webp","https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance-800.webp","https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance-600.webp","https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance-300.webp","https://static.dastra.eu/content/bcec9acf-a4e2-4f7b-9be2-217848ee54a1/ai-act-compliance-100.webp",61765,[]]