MEPs will vote on new rules to ensure artificial intelligence is trustworthy, safe and respects EU fundamental rights, while supporting innovation. MEPs debated on Tuesday 12.03 and approved on Wednesday an agreement reached with EU countries establishing obligations for AI based on its potential risk and level of impact. AI applications that pose a clear risk to fundamental rights - such as biometric categorisation systems based on sensitive characteristics, social scoring, or AI used to manipulate human behaviour - will be banned in Europe. AI systems considered high-risk, used for example in critical infrastructure, education, healthcare, law enforcement, border management or elections, will have to comply with strict requirements. Fast-expanding general-purpose AI (GPAI) models will also have to comply with transparency obligations and EU copyright rules, while the most powerful models will face additional safety requirements.

Strasbourg

debate

plenary session

VOSS, Axel (EPP, DE)

use of robotics and artificial intelligence (T)

BENIFEI, Brando (S&D, IT)

PLUMB, Rovana (S&D, RO, S&D, RO)

MAYDELL, Eva (EPP, BG)

SOLÍS PÉREZ, Susana (Renew, ES)

TUDORACHE, Dragoş (Renew, RO)

BRETON, Thierry (FR, EC)

Artificial Intelligence Act: extracts from the debate (12.03) and the vote

https://api.multimedia.europarl.europa.eu/documents/20125/75311534/1709890655256_I254339Chapters%5BSD-EN%5D.srt

SOUNDBITE (Italian), Brando BENIFEI (S&D, IT), Rapporteur, After two years of, hard work, we have now reached the approval of the first binding law on artificial intelligence. Reduce risks, to combat, discrimination, to give, transparency, to make sure that equality and freedom are sustained and, at the same time, supporting innovation and competitiveness.

SOUNDBITE (Italian), Brando BENIFEI (S&D, IT), Rapporteur, We are at the moment, the preparing an AI office, an office within Europe. Institutions and companies need to be helped for voluntary compliance in order to apply these rules before they are fully compulsory.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, The EU now has an AI office to govern the most powerful AI models. The US and the UK have established equivalent structures and many more will follow suit. It is imperative to connect these initiatives into a strategic triangle and then into a network. And that's what allow us to stay on top of any unforeseen developments in AI that might require a coordinated response.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, We need to put our entire diplomatic and political clout in promoting the European model of AI governance. This is a strategic priority for the next mandate, alongside reducing our strategic dependencies and increase our resilience. Because the future is AI fueled and we must continue to shape it.

SOUDNBITE (English), Eva MAYDELL, (EPP, BG), The way I see it is that we are laying out a common European vision for the future of this technology, one where AI is more democratic and safe, but also, I would hope so, more competitive.

SOUNDBITE (German), Axel VOSS (EPP, DE), Apart from the technical harmonization of standards, we need a flexible approach to all of the problems which may arise and we shouldn't believe, and the rapporteur said this, that our work is just going to end here with this legislation. I think it's only just starting.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, We've focused on two level approach transparency for all models of artificial intelligence for general use, and then stricter needs for more powerful models, which could have a potential systemic impact on our single market in the European Union.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, From where we stand today. The European Parliament, the Council and the Commission. We have shouldered our responsibilities. And these responsibilities are translated into the text, which you will be approving tomorrow by a large majority tomorrow.

SOUNDBITE (Spanish), Susana SOLÍS PÉREZ (Renew, ES), on behalf of the ENVI Committee, European citizens can rest easy knowing that our fundamental rights are going to be protected more. But lets not just talk about the risks today, this does actually offer benefits, this technology is for our society. 90% of its uses are low risk. Protection of citizens rights and promoting innovation need to be compatible in the European Union, because if we don't have that, we're going to drive innovators out of Europe. We have got a decent umbrella law. It's the beginning. As has been said, we need to approach it in the sector by sector weighted, flexible way so that companies can keep up with the developments of technology.

SOUNDBITE (English), Rovana PLUMB (S&D, RO), on behalf of the TRAN Committee, We prioritize the safety and privacy of public transport users, as well as flexibility for small and medium enterprises. Therefore, we warmly welcome the establishment of stringent conditions for the further processing of personal data pertaining to the safety and resilience of the transport systems and mobility within AI regulatory sandboxes.