10 Apr 2026 AI EngineeringUpdated 21 Aug 20266 min read

EU AI Act Article 50 Now Applies — What to Implement

Article 50 has applied since 2 August 2026. We distinguish provider and deployer duties, explain the limited grace period to 2 December 2026, and turn the final guidelines and voluntary Code of Practice into a practical checklist.

Article 50 Has Applied Since 2 August 2026

The transparency duties in Article 50 have applied since 2 August 2026. The deferred dates for high-risk AI — 2 December 2027 for stand-alone systems and 2 August 2028 for systems embedded in products — do not suspend those duties. The common shorthand that “the AI Act was postponed” is therefore wrong.

Article 50 requires concrete product changes: notices for direct AI interaction, machine-readable marking of certain generated outputs, and human-perceivable disclosure in specified deployment cases. A limited grace period to 2 December 2026 applies only to systems placed on the market before 2 August 2026 and only to the marking and detection duty in paragraph 2. What follows is technical guidance, not legal advice.

Untrusted input arrives — treat it as data, not instructions. 1/4

What Article 50 Actually Requires

Article 50 of Regulation (EU) 2024/1689 carries four substantive duties across two roles. Paragraphs 1 and 2 bind the provider — whoever places a system on the market under their own name. Paragraphs 3 and 4 bind the deployer — whoever uses it under their own authority. Paragraph 5 governs the form of the notice. Paragraph 6 confirms these duties are additive to Chapter III and other Union law.

The exemptions are narrow. Paragraph 1 does not apply where the AI nature is obvious to a reasonably well-informed, observant and circumspect person. Paragraph 2 does not apply where a system performs an assistive function for standard editing or does not substantially alter the input data or its semantics. Paragraph 4 reduces the duty for evidently artistic, creative, satirical or fictional works to a disclosure that does not hamper enjoyment.

Article 99(4)(g) caps penalties for breaches of Article 50 at EUR 15 million or 3 percent of total worldwide annual turnover, whichever is higher. Article 99(6) inverts that for SMEs: whichever is lower. Member States set the actual penalties and designate the authority that imposes them.

Provision	Bound party	Trigger	Duty
Art. 50(1)	Provider	System interacts directly with a person	Inform that this is an AI system
Art. 50(2)	Provider	Synthetic audio, image, video or text	Machine-readable mark, detectable output
Art. 50(3)	Deployer	Emotion recognition, biometric categorisation	Inform the persons exposed
Art. 50(4)	Deployer	Deepfake image, audio or video	Disclose artificial generation
Art. 50(4)	Deployer	AI text on matters of public interest	Disclose, unless editorial responsibility
Art. 50(5)	Both	All of the above	Clear, distinguishable, at first exposure, accessible

Marking Is a Provider Duty

Article 50(2) requires that the outputs of generative systems are marked in a machine-readable format and detectable as artificially generated or manipulated. It names no technique. It sets a standard of care instead: effective, interoperable, robust and reliable as far as technically feasible, taking into account the state of the art and the cost of implementation. The duty sits with the provider, including providers of general-purpose models offered as a service.

The final Code of Practice was published on 10 June 2026. It is voluntary: signatories may use its measures as a predictable route to demonstrate compliance with paragraphs 2, 4 and 5; non-signatories must demonstrate compliance through other adequate means. The Code does not replace the legal text.

A technical implementation can combine signed provenance metadata, a signal watermark, and a provider-side generation log. That combination is an engineering recommendation, not the literal wording of Article 50. Every layer has limits: upload pipelines can strip metadata, editing can defeat watermarks, and a log is only as reliable as its access controls, key management, and retention.

Disclosure Is an Interface Problem

Paragraph 5 is the requirement teams underestimate. The information must be clear and distinguishable, must reach the person at the latest at the time of first interaction or exposure, and must meet applicable accessibility requirements. A dismissible modal at signup satisfies none of that for a widget on a third-party page, a deep link into an existing conversation, or a voice channel where nothing is visible.

We build disclosure as a property of the channel rather than of the session. Every surface that renders model output declares its own notice: a persistent label in chat, a spoken sentence before the first turn in voice, an explicit line in email and ticket replies. The notice reaches assistive technology as text, not as colour or an icon alone. Under paragraph 3 the notice comes before processing starts.

Provider or Deployer Changes Everything

Duties follow the role, and the role is not decided by who wrote the code. A provider places a system on the market under its own name or trademark. A deployer uses it under its own authority. Under Article 25, a deployer that puts its own name on a system, changes its intended purpose, or makes a substantial modification becomes a provider of that system — and inherits paragraph 2 with it.

Most organisations we work with are deployers of a general-purpose model and providers of the application built on it. That combination is the common failure mode. The vendor marks what it generates; the application then composes, crops, converts and re-renders that output, and the mark does not survive the pipeline. Marking must be re-applied where your system emits the artefact. Put that expectation in the upstream contract and verify it in a test.

What the Act Does Not Require

Article 50 does not require a visible watermark on every generated image. Provider-side machine-readable marking and human-perceivable deployer disclosure are separate duties. The final guidelines exclude, among other things, source code, very short character sequences, purely machine-to-machine output, and certain non-final output in closed development environments. “Not published” is not a blanket exemption.

It also guarantees nothing. A mark shows that a system claims to have generated a piece of content. It does not show that unmarked content is human, and detection tools do not remove that asymmetry. Article 50 is a disclosure duty on identified operators, not a provenance system for the open web. Compliance says nothing about the GDPR, the DSA or copyright, which apply in parallel under paragraph 6.

What Moved — and What Did Not

Regulation (EU) 2026/1744, the Digital Omnibus on AI, was published in the Official Journal on 24 July 2026 and entered into force on 27 July. Its fixed timeline moves stand-alone high-risk AI rules to 2 December 2027 and product-embedded high-risk AI rules to 2 August 2028. Article 50 has nevertheless applied since 2 August 2026.

The only transition relevant here is narrow: systems already placed on the market before 2 August 2026 need to meet paragraph 2 only from 2 December 2026. It does not defer the notice for direct AI interaction or the deployer duties concerning emotion recognition, biometric categorisation, deepfakes, or specified public-interest text.

National market surveillance authorities carry most enforcement. The AI Office supervises only specified systems based on general-purpose models and certain very large platforms or search engines. That division does not change the duties of the provider or deployer in scope.

Build the Marking Layer Now

The principal implementation material is now final: the voluntary Code of Practice since 10 June 2026 and the Commission guidelines since 20 July 2026. The guidelines clarify roles, exceptions, and examples; the Code provides a voluntary compliance route. Teams should no longer implement against drafts.

We therefore build on what is already fixed: a defined point in the architecture where generated artefacts are marked and logged; a disclosure component owned by the channel; and a labelling decision recorded in the publishing pipeline together with the person who made it. Signing keys, rotation and verification are treated as production infrastructure.

Signed identities and generation logs can also improve operational traceability. That is a possible additional benefit, not a security effect established by Article 50: a mark does not prove that unmarked content is human or that marked content is true.