The European Parliament has approved regulations meant to set a worldwide benchmark for the technology, which covers a broad spectrum from automated medical diagnoses to certain types of drones, AI-produced videos referred to as “deepfakes”, and bots such as ChatGPT.

The Irish Council for Civil Liberties (ICCL) said the text approved by MEPs includes a complete ban on the use of real-time facial recognition technology (FRT) in public spaces in a “significant blow” to the Irish government’s plans to introduce FRT for An Garda Síochána.

MEPs will now negotiate the details with EU countries prior to the draft rules becoming law in the form of an AI Act.

Dr Kris Shrishak, senior fellow at ICCL, said: “The EU’s AI Act is likely to become a global standard for AI regulation and so the stakes for people’s fundamental rights are extremely high. Today’s vote sends a clear message to the other EU institutions that the Parliament feels AI should be strictly regulated to protect EU citizens’ rights.

“The Parliament addresses some of the risks from AI systems like ChatGPT, which is welcome. But we still have many concerns. The Parliament should take a firm stand during the trialogues to prevent companies from gobbling up data for free and using copyright-protected data for training without consent.

“The Parliament has also tied regulators’ hands by preventing them from accessing the source code of AI systems for investigations. We are concerned that, similar to the GDPR, we will see problems when it comes to enforcing this legislation.”

Data protection lawyers said the EU was set to diverge significantly from the UK’s regulatory framework for AI.

Claire Edwards, a Manchester-based partner at Addleshaw Goddard, said: “The UK seems to be taking a much more hands-off approach to regulation of AI. A lot of AI companies see the UK as a welcoming environment and like that there’s already a skilled workforce in place for them to recruit from.

“Clearly for those who want to licence solutions into the EU or which would offer services in the EU then the AI Act is relevant regardless of their physical corporate location.

“Other countries are developing their own laws and frameworks on AI, including the UK, USA and China, and there are various proposals for international frameworks. It is likely that lawmakers and other organisations working on these will pay close attention to the draft EU Act and assess its strengths and weaknesses.

“I think the EU recognises both the major potential and major risks associated with AI. The bloc has chosen to take a coordinated approach to avoid member states creating domestic legislation that is inconsistent with that of neighbouring countries.

“The potential implications appear to be that AI will be more heavily regulated in the EU than in the UK. While the EU approach seeks to balance the risks by categorising AI tools into risk levels, some critics believe that it will stifle innovation.

“Potentially, as AI develops further, the EU may not be given access to various platforms as a result of this legislation.”