Responsible AI Platform

AI Act Monitor

What changed, and what affects you?

Material changes in legislation, guidance, standards and supervision, reduced to their practical meaning.

Editorially verified on How we monitor

Make the monitor relevant to you

Make this relevant to youOptional
For whom?
Topics
Choose one or more topics

7 relevant changes

RSS

Action or reassessment

4
ApplicableAction required

Article 50 applies, with final guidance on its scope

What changed

The Commission adopted final guidelines on Article 50 on 20 July 2026 and updated the related questions and answers on 24 July.

Possible next step

Inventory chatbots, generative features, deepfakes, emotion recognition and biometric categorisation separately.

Relevant for: Using AI, Providing AI, Advising +2Topics: Transparency, Supervision and enforcement

ApplicableAction required

Enforcement of GPAI obligations has started

What changed

Since 2 August 2026, the Commission can enforce the obligations for providers of general-purpose AI models. The substantive obligations have applied to new models since 2 August 2025.

Possible next step

Record whether the organisation is a general-purpose AI provider, a downstream provider or only a deployer.

Relevant for: Providing AI models, Providing AI, Using AI +2Topics: GPAI, Supervision and enforcement, Transparency

Final guidanceReassessment required

Article 4 is now clearly a duty to take measures

What changed

The Commission updated its questions and answers on AI literacy on 27 July 2026 after Regulation (EU) 2026/1744 entered into force.

Possible next step

Remove claims about a mandatory individual certificate, test result or minimum number of training hours.

Relevant for: Using AI, Providing AI, Advising +2Topics: AI literacy, High-risk AI

Binding and in forceReassessment required

The Digital Omnibus has become binding law

What changed

Regulation (EU) 2026/1744 was published on 24 July 2026 and entered into force on 27 July 2026. It directly amends the AI Act.

Possible next step

Replace the former high-risk dates in policies, registers and project plans.

Relevant for: Using AI, Providing AI, Providing AI models +3Topics: AI literacy, High-risk AI, FRIA, Supervision and enforcement

Important developments

3
Published standardImportant development

First AI Act quality management standard has been published

What changed

EN 18286:2026 has been published as a European quality management system standard for regulatory purposes under the AI Act. NEN has adopted the standard nationally.

Possible next step

Compare the existing quality management system with EN 18286:2026.

Relevant for: Providing AI, Using AI, Advising +1Topics: Standards, High-risk AI

Draft guidanceImportant development

Draft provides examples for high-risk classification

What changed

On 19 May 2026, the Commission published three draft guidelines on the classification of high-risk AI. The consultation has closed, but the guidelines have not yet been finally adopted.

Possible next step

Provisionally test existing classifications against the Annex I and Annex III examples.

Relevant for: Using AI, Providing AI, Advising +1Topics: High-risk AI, FRIA

Only when something changes

Follow all material changes

Receive only material changes that match this selection. No more than once a week.