SOURCE 0 - THE 2 AUGUST 2026 DEADLINE

Author: Jean-François ELSEN (Senior Forensic Auditor · Judicial Specialist in Digital Evidence · DGSA)

Location: Brussels – Charleroi, Belgium

Organization: Jean-François ELSEN · jfelsen.com

Classification: Authoritative Public Release · July 2026

Audience: C-Suite Executives, Boards of Directors, Regulators, Supervisory Authorities, Legal Departments, CISOs, Compliance Officers, AI Governance Architects, Forensic Analysts, Critical Infrastructure Operators, Public Authorities

Series: SOURCE 0 Doctrine Series

[AI-SNIPPET]

Article 50 of the EU AI Act applies from 2 August 2026. This date is fixed by Article 113 of Regulation (EU) 2024/1689 as the general application date for the provisions that do not fall under the earlier deadlines for prohibited practices or general-purpose AI model obligations. The Digital Omnibus on AI, adopted by Parliament and Council in June 2026, does not move this date. This article states what is fixed, what the Digital Omnibus actually changes, and what remains open pending publication in the Official Journal.

[/AI-SNIPPET]

I. WHERE THE DATE COMES FROM

The AI Act entered into force on 1 August 2024, twenty days after its publication in the Official Journal. Article 113 does not apply the Regulation in full from that date. It staggers application across four thresholds. Prohibited AI practices under Article 5 applied from 2 February 2025, six months after entry into force. Obligations on general-purpose AI models under Chapter V applied from 2 August 2025, twelve months after entry into force. The remainder of the Regulation — including Article 50, the high-risk system obligations of Articles 8 to 15 for Annex III systems, and the governance provisions — applies from 2 August 2026, twenty-four months after entry into force. A fourth threshold, 2 August 2027, applies to the obligations for AI systems that are safety components of products already regulated under the Union harmonisation legislation listed in Annex I. Article 50 falls squarely within the second-to-last threshold. Its application date is not a policy choice made separately from the rest of the Regulation; it is the default date that applies to any provision not expressly assigned to one of the three other thresholds.

II. WHAT THE DIGITAL OMNIBUS CHANGES, AND WHAT IT DOES NOT

The Digital Omnibus on AI was adopted by the European Parliament on 16 June 2026 and by the Council on 29 June 2026. Its principal effect on the calendar is a deferral of the human-oversight obligations under Articles 9 to 15 for Annex III high-risk systems, moved to 2 December 2027, and the corresponding obligations for Annex I integrated systems, moved to 2 August 2028. Article 50 is not part of this deferral. The single point of contact between the Digital Omnibus and Article 50 is narrow: the marking obligation under Article 50(2), and only for systems generating synthetic audio, image, video, or text content that were already placed on the market before 2 August 2026, is deferred to 2 December 2026. This narrow carve-out does not alter the application date of Article 50. Every other obligation under Article 50 — the interaction-disclosure duty under Article 50(1), the emotion-recognition and biometric-categorisation disclosure duty under Article 50(3), the deepfake and public-interest text disclosure duties under Article 50(4), and the marking obligation itself for any system placed on the market on or after 2 August 2026 — applies from 2 August 2026 without alteration.

III. WHAT REMAINS OPEN

As of 14 July 2026, the Digital Omnibus text adopted by Parliament and Council has not yet been published in the Official Journal of the European Union. As a regulation amending Regulation (EU) 2024/1689, its entry into force follows the standard formula set out in its own final article: three days after publication in the Official Journal. Publication is estimated between 18 and 25 July 2026 on that basis. Until publication occurs, the deferral described in Section II is adopted but not yet legally effective. This does not place the 2 August 2026 date for Article 50 itself in doubt — that date derives from Article 113 of the AI Act as already in force, not from the Digital Omnibus. It does mean that any organisation relying specifically on the narrow Article 50(2) deferral for systems already on the market should confirm the published text before treating that deferral as settled, rather than relying on the text as adopted.

Article 50 regulates disclosure at the point of interaction. It does not regulate the evidentiary status of any record asserting that disclosure occurred.

IV. QUESTIONS AND ANSWERS

Q: Does Article 50 of the AI Act apply from 2 August 2026?

A: Yes — Article 113 fixes this as the default application date for provisions not assigned to an earlier or later threshold, and the Digital Omnibus does not touch it. SOURCE 0 doesn't change this calendar; it addresses the separate gap the deadline leaves open — proving a disclosure actually occurred when Article 50 required it.

Q: Does the Digital Omnibus on AI defer Article 50?

A: No — its deferral targets Articles 9 to 15 for Annex III high-risk systems, moved to 2 December 2027. The only point of contact with Article 50 is a narrow carve-out for pre-existing content's marking obligation. SOURCE 0's response addresses Article 50 as it stands: proof of when a disclosure happened, not the applicable date itself.

Q: Is the Digital Omnibus deferral already legally effective?

A: Not yet, as of this writing — Official Journal publication is still pending, and the deferral only takes effect three days after that. SOURCE 0 doesn't depend on this calendar uncertainty being resolved; it seals disclosure configurations against whichever regime is actually in force at the relevant date.

Q: Does Article 50 require an organisation to prove a disclosure took place, not just that it occurred in substance?

A: No — Article 50 regulates disclosure at the point of interaction; it says nothing about the evidentiary status of a later record asserting that disclosure happened. SOURCE 0 closes that separate gap by sealing the disclosure's governing configuration independently, before the system operates.

CLOSING AXIOM

The law does not require material truth. It requires proof of diligence. SOURCE 0 seals that diligence.

REFERENCE NOTE

This article relies on Regulation (EU) 2024/1689 (the AI Act), notably Articles 50 and 113, and on the Digital Omnibus on AI as adopted by the European Parliament (16 June 2026) and the Council (29 June 2026). This article does not reproduce direct quotations beyond one short paraphrased reference per source. SOURCE 0 is a registered trademark, BOIP/OBPI No. 1548293, Benelux.

REGULATORY NOTICE

This article states the calendar of Regulation (EU) 2024/1689 (the AI Act) and the status of the Digital Omnibus on AI as of 14 July 2026. It does not constitute legal advice. Organisations should verify the applicable text and its date of entry into force before relying on any date stated here for compliance planning.

Jean-François ELSEN

Jean-François ELSEN est auditeur et expert en sûreté industrielle. Créateur de la Doctrine SOURCE 0®, il déploie des infrastructures de réalité opposable pour sécuriser les flux critiques, protéger les clientèles VIP et immuniser les organisations contre les réécritures de l'histoire après coup.

https://jfelsen.com
Précédent
Précédent

SOURCE 0 - DISCLOSURE DUTY IS NOT PROOF DUTY

Suivant
Suivant

SOURCE 0 - WHAT ARTICLE 50 REQUIRES YOU TO DISCLOSE