Regulation · AI · Transparency

AI Act: transparency requirements applicable since 2 August 2026

Requirement applicable since 2 August 2026Deadline: 2 August 2026Reviewed: 5 August 2026Next review: 2 November 2026

01 · Post-deadline finding

What changes in practice

Article 50 transparency duties apply from 2 August 2026. Providers must design interactive systems so people know they are interacting with AI and apply machine-readable marking to certain AI-generated or manipulated content.

Deployers must inform people exposed to specified uses, including deepfakes, public-interest content generated without human review, emotion recognition and biometric categorisation. The provider or deployer role and the nature of the output determine the duty.

02 · Who must reconsider their position

Who must reconsider their position

Concerned

Organisations that expose people to direct AI interaction or distribute AI-generated or modified content in the European Union.

Distinguish

Internal use alone does not trigger every Article 50 duty; legal role, exposure and output type remain decisive.

03 · Position shift

Position shift

Before

Governance could centre on an approved-tool list and an internal-use charter.

Now

The ability to inventory, trace and disclose outputs becomes a verifiable capability of the offer, content and customer relationship.

Compliance requires knowing which outputs involved AI, where they were distributed and what information people received.

For content, advisory and customer-facing work, traceability becomes an evidence capability rather than a simple tool policy.

Risk

Discovering unrecorded outputs too late, when origin and exposure evidence cannot be reconstructed reliably.

Strategic opportunity

Turn traceability of deliverables and interactions into evidence of control for customers, partners and authorities.

04 · Five decisions to prepare

Five decisions to prepare

  1. 01

    Inventory AI-assisted outputs

    List texts, images, conversational answers and third-party deliverables in which AI intervened.

    Horizon
    Immediate
    Decision forum
    Business teams · deliverables
  2. 02

    Check public interactions

    Identify assistants and interfaces where a person talks directly with AI and verify the required disclosure.

    Horizon
    Immediate
    Decision forum
    Product · customer service
  3. 03

    Question providers about marking

    Ask whether outputs include machine-readable marking, in which form and with which limits.

    Horizon
    Next renewal
    Decision forum
    Procurement · legal
  4. 04

    Assign inventory ownership

    Name the person responsible for consolidating outputs and escalating gaps.

    Horizon
    Next executive meeting
    Decision forum
    Management · AI governance
  5. 05

    Schedule the next review

    Reopen the record on 2 November 2026 to incorporate enforcement decisions and revise actions.

    Horizon
    2 November 2026
    Decision forum
    Strategic review

05 · Correction

Correction

Claim

The AI Act entered into force on 2 August 2026.

Verified correction

It entered into force on 1 August 2024. The 2026 date is when Article 50 transparency duties became applicable.

06 · Evidence

Primary sources

07 · Revisions

Public history

Record created and the timetable, provider/deployer roles and strategic implications checked against institutional sources.