cross-posted from: https://scribe.disroot.org/post/3556484

For Ireland, which will soon take over the Presidency of the European Council, the AI Act is an opportunity to wipe the slate clean after a decade of tax, tech and data enforcement scandals. The State will concentrate on AI during its EU Presidency next year.[1] Yet, Ireland’s poor tech enforcement record is likely to continue with AI unless the State takes immediate action.

On 3 July, ICCL Enforce wrote to the Minister for Enterprise, Tourism and Employment urging him to rapidly designate the responsible “market surveillance authorities” (MSAs), the regulators responsible to enforce the prohibitions under the AI Act.[2] This is urgent because these authorities will be responsible for enforcing prohibitions on dangerous AI in just a few days from now (2 August 2025).[3]

When asked in a Parliamentary Question, the Minister was unable to name the MSAs responsible to enforce the prohibitions last week.[4] This means there no enforcer knows whether it is responsible for monitoring the Tiktok recommender systems that significantly harm children.

Ireland also needs to designate regulators for “high-risk” uses of AI by 2 August. So far, the State has named just one regulator, the Data Protection Commission, which is responsible for three of the eight high-risk AI uses. For all others, including education, critical infrastructure and access to essential services, the Minister could say no more than “arrangements to be finalised”. [5]

The State is required by EU law to provide regulators with “adequate technical, financial and human resources, and with infrastructure”[6] by 2 August. It cannot do so if it does not know who those regulators will be.

[…]