MEPs debate on new rules to ensure artificial intelligence is trustworthy, safe and respects EU fundamental rights, while supporting innovation. MEPs are expected to approve tomorrow 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. Opening statements by Brando BENIFEI (S&D, IT) and Dragoş TUDORACHE (Renew, RO), Rapporteurs, Thierry BRETON, Commissioner for Internal Market, Axel VOSS (EPP, DE), on behalf of JURI Committee, Susana SOLÍS PÉREZ (Renew, ES), on behalf of the ENVI Committee, and Rovana PLUMB (S&D, RO), on behalf of the TRAN Committee. This transcription is a result of an automatic process. Please check against delivery. MAYDELL Eva

statement

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)

KOLAJA, Marcel (Greens/EFA, CZ)

SOLÍS PÉREZ, Susana (Renew, ES)

TUDORACHE, Dragoş (Renew, RO)

BRETON, Thierry (FR, EC)

Artificial Intelligence Act: Opening statements by Brando BENIFEI (S&D, IT) and Dragoş TUDORACHE (Renew, RO), Rapporteurs, and Thierry BRETON, Commissioner for Internal Market

SOUNDBITE (Italian), Brando BENIFEI (S&amp;D, IT), Rapporteur, Thank you, President, Commissioner. Well, after two years of hard work, we have now reached the approval of the first binding law on artificial intelligence to 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. Thanks to us, to the European Parliament.

SOUNDBITE (Italian), Brando BENIFEI (S&D, IT), Rapporteur, After hard negotiations, practices that are unacceptable in Europe will be banned. For instance, the recognition of emotions at the workplace or at a place of study, or to define the religious or ethnic relevance of biometric on the basis of biometric systems. The workers must also always be informed too. We must be told of the use of artificial intelligence at the workplace. We will be protecting the rights of citizens and workers. And these are some of the important examples of what we have achieved.

SOUNDBITE (Italian), Brando BENIFEI (S&D, IT), Rapporteur, In the final text, we have also been able to reduce the risk in sensitive areas. For instance, the use in medical activities. We need to verify the conformity on the reliability of the data, digital security. So, we put obstacles to artificial intelligence, which tries to discriminate women or people of other ethnic origins. These are data that we are familiar with and we wanted to take a law that can have an effect here in this area. We also want to make sure that the we want to create an incentive for the use of artificial intelligence. And people can use systems but with a maximum guarantee.

SOUNDBITE (Italian), Brando BENIFEI (S&D, IT), Rapporteur, And thanks to the work of the European Parliament, we've also worked on the basis of transparencies. So, we must make sure that big fakes are no longer use. They are recognised by the devices, by warfare marking, which is a kind of architecture which is invisible. We will also put an end to use of those issues protected by copyright, because we understand the economic value of creativeness and this is a right which needs to be protected within the context of artificial intelligence. There is a lot that remains to be done.

SOUNDBITE (Italian), Brando BENIFEI (S&D, IT), Rapporteur, In order to improve this, we are at the moment the preparing and AI Office within Europe. Institutions and companies need to be helped for voluntary compliance in order to apply these rules before they are fully compulsory. It will be very gradual and it will put at the very centre, for instance, the development of what we wanted in the Parliament. That is to verify fundamental rights, verify on impact. This is fundamental, so that the uses of artificial intelligence can also play their role to reduce risks.

SOUNDBITE (Italian), Brando BENIFEI (S&D, IT), Rapporteur, 'Allow me to thank all of us for this very important task, which is putting Europe at the very forefront of this particular type of legislation. And we will have to work especially on artificial intelligence and at the workplace. So, we have done this work together with the Co-rapporteurs, our colleague, Tudorache and all the other colleagues and shadow rapporteurs and the political groups. The Parliament's team has been united when negotiating in the light of our views. That is to truly put the human being at the very centre of the development of artificial intelligence. Thank you very much.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, Thank you very much, Chair, Commissioner. Dear colleagues, it is customary to begin with thanks. I will actually continue where my colleague left it, and I want to say first of all, thank you, Commissioner Breton, and thank you to the Commission as such for the foresight of putting forward this proposal with no one else around the world was actually looking at this option. I also want to thank the shadow rapporteurs for putting the incredible energy and effort that was required. From you and your teams to draft those over 3000 amendments to streamline them into a coherent mandate and then negotiate that with the Council.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, 'I would think the Council, if they were here, but they are not. And I want to specifically thank Brando Benifei for being such a great colleague and co-rapporteur. So this is it, colleagues, we delivered. The EU delivered. No ifs, no buts, no later. These are the rules. We have forever attached to the concept of Artificial Intelligence, the fundamental values that form the basis of our societies. And with it alone, the Act has nudged the future of AI in a human centric direction and a direction where humans are in control of the technology and where it the technology helps us leverage new discoveries, economic growth, societal progress and unlock human potential.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, Much work lies ahead that goes beyond the act itself. I will push us to rethink the social contract resting at the heart of our democracies, along with our education models, our labour markets, the way we conduct warfare. All this to say that the act is not the end of a journey, but rather the starting point for a new model of governance built around technology. Going back to the Act itself, we must now focus our political energy in turning it from the law in the books to the reality on the ground.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, First, implementation, the rules we have passed in this mandate to govern the digital domain, not just the act, are truly historical pioneering, but making them all work in harmony with a desired effect and turning Europe into the digital powerhouse of the future will be the test of our lifetime. Second, we need to work with others. The EU now has an office to govern the most powerful models. The US and the UK have established the equivalent structures and many more will follow soon. 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 a way that might require a coordinated response.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, Third, we need to play on the global scene in accordance with our weight, not just by making the rules, but also by exporting them. We need to put our entire diplomatic and political clout in promoting the European model of governance. This is a strategic priority for the next mandate, alongside reducing our strategic dependencies and increase our resilience because the future is fuelled and we must continue to shape it. One final thought. We haven't yet witnessed the full power of the AI. Existing capabilities have already surpassed anything we could have imagined a decade ago, at speeds we have never thought possible. And this will continue exponentially.

SOUNDBITE (English), Dragoş TUDORACHE (Renew, RO), Rapporteur, For lack of a better term, Artificial General Intelligence or AGI, is something we need to prepare for. And AI possessing an intelligence that is superior to that of a human being will open up infinite possibilities but at the same time will raise never before encountered ethical, moral and yes, even existential questions. We have in the AI Act a answer first to those questions. But we must be prepared to cater our governance for quantum leaps that we know AI could do. I'll conclude not as a politician, but as a parent. This regulation makes me feel more confident about the future of my children, and I am humbled to have played a part in shaping it together with you all those present in this Chamber. Thank you.

SOUDNBITE (English), Eva MAYDELL, (EPP, BG), on behalf of the ITRE Committee, Good afternoon colleagues, Commissioner, we've arrived to the week that we are going to have the final vote of the first ever law on Artificial Intelligence. 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. That is, if it's done, right. In a way, we close one chapter, but the story continues. And what I believe now that it's important is that we focus on three things.

SOUDNBITE (English), Eva MAYDELL, (EPP, BG), on behalf of the ITRE Committee, First of all, to provide legal certainty to all stakeholders, including companies. Second of all, to play an active role in ongoing global efforts through the G7 process, OECD and the UN. And thirdly, to prioritise AI excellence through concrete commitments and investments. I stand before you because I do believe that in Europe we are more than just a referee. I believe we are a continent of innovators, of doers, of dreamers, of achievers. And now it's time for us to walk that talk when it comes particularly to promoting AI excellence in Europe. Thank you.

SOUNDBITE (German), Axel VOSS (EPP, DE), on behalf of JURI Committee, President, Commissioner, colleagues. At the end of the session that we have before us is a complex, it's a complicated and I thank for leading IA and it would be important. As the atmosphere outside the EU bubble is not particularly good, so I'm missing some added value in terms of innovation. We are not really leading the pack in the EU. So, we believe that the IA law is only sensible if we take further measures and add points to ensure that this is a success. Apart from the technical harmonisation of standards, we need a flexible approach to all of these 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.

SOUNDBITE (German), Axel VOSS (EPP, DE), on behalf of JURI Committee, I think it's only just starting. We need to respond more quickly than we usually do and adapt to problems and new developments. We need to avoid legal uncertainty due to a lack of clear wording in the text. So we need to clarify the IA Office and have guidelines as quickly as possible. We need to determine where there is overlap with other legislation and minimise that and clarify. We need an interpretation of the regulations. So, for the entire European Union, a single uniform interpretation. We need a concept for the sandboxes, so that we allow a certain sandbox, a specific sandboxes for certain sectors, and we need to to ensure that we have adaptation to the GPT,  in terms of personal data. And SMEs need to be provided with extreme assistance in order to overcome the challenges. And with regard to the office, we need to ensure it has the right talent that is also essential. Thank you.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, Thank you, President. Ladies and gentlemen, Honourable Members of the European Parliament. And I'm also turning to the audience. There's a lot of people up in the audience gallery, and I see a lot of young people, and I believe that this is inspiring, inspiring for us and inspiring for you to be here today and to attend this particular debate, which somehow closes an historic cycle, as was already said by the two Rapporteurs, that I'd like to greet, dear Brando, dear Dragoş.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, Yes, it is an historic debate which is being held here and has been held for several months to lead to tomorrow's vote, a vote which will make it possible for us, as has already been said by the rapporteurs, to develop the first regulation on artificial intelligence in the world, which is a balanced text which is there to protect us from any excesses that may occur to the negligent use of artificial intelligence, but also to promote innovation.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, On the 8th of December, it feels almost like it was yesterday, Dragoş and Brando. We spent 38 hours, 38 hours together. And I'll tell the audience, this is the longest negotiation that was ever held within the framework of the trialogues in the history of our European institutions. So, this obviously clearly indicates the subject was very complex and it was necessary to take time to harmonise the position of the European Parliament, the legitimate views of the Parliament and Council. And I'd like to welcome the well work that was done by the Minister, the Spanish Minister, Carme Artigas, during the Spanish presidency.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, And we reached an historic compromise. It is balanced and it will resist the passing of time. Allow me to focus on four aspect particularly. Together, we managed to find the fair balance for risk management and promotion of innovation. Allow me to remind you that it is because of risk that we are regulating. This approach based on risks, bonds in Europe, the use of artificial intelligence which violates fundamental rights and values. And it is also fixing rules in high risk cases. In the case of health issues, for instance, and also to promote innovation with no barriers, no obstacles for low risk use. And that is obviously the largest number of cases.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, Secondly, we also developed a balanced, restrictive framework for the Artificial Intelligent models for general use. And this is an initiative that was taken by the European Parliament, by the rapporteurs, by the shadow rapporteurs, and I would like to thank you for your commitment. This is legislation on artificial intelligence without measures of large artificial intelligence systems for general use would not have been efficient. So, we focussed 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. And this way what we've managed to avoid is excessive load on the innovators and more transparency. As was already said, for many SMEs, which have always been at the heart of our concerns.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, Thirdly, we have a solid framework for transparency for the generation of content. Think about manipulated images on the web. Those texts generated by artificial intelligence and published to inform as public information have been generated artificially. And we need to indicate that this is so. What about deep fake, deep fake artificially generated must be clearly indicated as being developed by artificial intelligence. And then we are also talking about a legal framework, law enforcement and this makes AI act very different from all the other issues that exist in other countries.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, Because it's a tough job to actually reach this type of agreement, and it leads to surveillance of market at national level. And it's also been said by the rapporteurs, it leads to a new office, European Office for Artificial Intelligence, which will be set up in the Commission within my services. We have already started recruitment and we have also set up a pact on artificial intelligence to make sure that hopefully tomorrow will be voting in favour of this particular resolution.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, And 350 organisations are working on this. And the main provisions for this Artificial Intelligence. So these measures will contribute to protect elections, many elections that will be taking place this year. But regulation cannot be the only thing that we do in order to strengthen leadership. In January, we approved a package in order to get the necessary ingredients for the development of artificial intelligence. No data, no artificial intelligence. The data sets a necessary V Power of calculations. These need to be very powerful. But we use this and this is the last point on the basis of algorithms, and that is the object of our engineers, of our scientists.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, 'And I know that they have a very good reputation in Europe. Europe already has three super calculators and two new ones will be developed. There will be in the top five, apart from these super calculators on the net, we should also be building AI factories where we will have super calculators focusing on data centres and services to support start-ups, so that they can benefit from this infrastructure to develop their innovations.

SOUNDBITE (French), Thierry BRETON, Commissioner for Internal Market, Ladies and gentlemen, Honourable Members. Tomorrow's vote will conclude 5 years of hard work on this particular act. We have had discussions, we have had exchanges, a few negotiations, debates, then interaction with the lobbies, all those who wanted to speak. We have been listening to them. But 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, I hope. Thank you. Commissioner.

SOUNDBITE (Spanish), Susana SOLÍS PÉREZ (Renew, ES), on behalf of the ENVI Committee, Thank you, Commissioner and President. We are giving today birth to a law on AI, which is unique. It looks at risk for people and it talks about penalties for failing to comply. It's going to make let us know if a text or photos have been artificially generated and things like copyright can be better protected in this way. You, European citizens, can rest easy knowing that our fundamental rights are going to be protected more. But it's not just talk about the risks today. This does actually offer benefits.

SOUNDBITE (Spanish), Susana SOLÍS PÉREZ (Renew, ES), on behalf of the ENVI Committee, This technology for our society. And 80% 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've got a decent umbrella law. It's the beginning, as it has been said. We need to approach it in a sector by sector weighted, flexible way so that companies can keep up with the developments of technology without deregulating. We need to seize the benefits of artificial intelligence and not be enslaved because Europe can have a big lead as far as AI is concerned, it's important for our competitiveness that we do. Thank you.

SOUNDBITE (English), Rovana PLUMB (S&D, RO), on behalf of the TRAN Committee, Thank chair. your Commissioner. Dear colleagues, tomorrow IA Act's adoption is pivotal for European policy making as it guarantees our rights and foster innovation.  In TRAN Committee. We prioritise 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 transport systems and mobility within AI regulatory sandboxes.

SOUNDBITE (English), Rovana PLUMB (S&D, RO), on behalf of the TRAN Committee, Moreover, granting priority access to regulatory sandboxes at minimal or proportionate cost for SMEs, and aiming to minimise the administrative burden and compliance costs demonstrates our commitment to their success. The AI Act will facilitate artificial intelligence development in sectors that directly bolster our citizens quality of life, especially in sectors like transport. I would like to congratulate the rapporteurs and all the shadows and Commission for this important legislative act. Thank you.