The AI Advisory Council has published its recommendations to the government on the alleged creation and public dissemination of AI-generated non-consensual intimate images, including child sexual abuse material (CSAM).

They come in response to an international scandal around images generated using Elon Musk’s AI chatbot Grok and shared on his social media platform, X — though the paper does not seek to establish specific facts in this case.

Irish law “is sufficiently robust to address the non-consensual sharing of AI generated intimate images and the production and/or sharing of AI generated CSAM”, according to the AI Advisory Council.

However, it says the “most effective response to the onset of technologies which allow for the production and publication of intimate images and CSAM at scale is a harmonised EU-wide approach”.

It notes that the EU AI Act “does not address AI technologies being used in this way” and suggests that this could be changed.

“The AI Advisory Council is of the view that the Irish government should use its assumption of the EU Presidency in the latter half of 2026 to work with the other EU member states and encourage the amendment of Article 5 of the AI Act pursuant to the Article 112(1) mechanism, in order to prohibit AI practices which permit users to generate non-consensual intimate images and CSAM,” the paper states.

It also calls on the Irish government to create a national taxonomy of online harms, with an explicit “AI-enabled and automated harm” layer, to standardise reporting and policy.

The paper’s two other recommendations relate to supporting victim reporting and evidence preservation, and organising an information campaign for the public.

The AI Advisory Council was established by the government in January 2024 and brings together leading legal and technological experts on AI.