Thierry Breton: “We will make full use of our toolbox to protect our citizens and our democracies”

 Three months before the European elections, Thierry Breton the EU Commissioner for Internal Market explains to “To Vima” that the EU “will make full use of our toolbox to protect our citizens and our democracies”. Mr Breton stresses that “in Europe, we now have the DSA (Digital Services Act) and we will make sure that platforms play by the rules. In addition, the AI Act will contain strict requirements for generative AI systems to mark artificially generated content”. Regarding AI and start-ups the Commissioner underlined that “we are making sure that European start-ups have access to the key ingredients of AI: data, computing power, algorithms and talent. Thanks to our data legislation and European data spaces, start-ups will have access to large volumes of industrial data to develop new AI applications”. As it has to do with his next steps, Thierry Breton said that “as we say: there’s no time but the right time! And for the time being, I am fully focused on my mission as Commissioner”. 

The AI Act is the world’s first comprehensive AI law. What does this mean for EU and in which fields could AI contribute?

The agreement on the AI Act makes Europe the first continent with a comprehensive regulatory framework for trustworthy AI. It follows five years of preparation, consultation and negotiation, resulting in a balanced set of rules that protects fundamental rights and promotes innovation in equal measure. It will allow our citizens and businesses to use AI safely and confidently, while giving AI developers the benefit of stable, harmonised rules across the single market.

The Regulation prohibits AI use cases that pose unacceptable risks (such as social scoring) and sets clear rules for high-risk use cases. At the same time, it sets out certain transparency measures for large language models that can be used for many different purposes – so-called general-purpose AI. If large, powerful AI models have a systemic impact in our single market, we expect those models to take certain risk management measures. Overall, this set of proportionate rules will allow Europe to be at the forefront of trustworthy AI.

I would also like to emphasise that the AI Act is pro-innovation thanks to its risk-based approach. We are regulating as little as possible, as much as needed. So-called “regulatory sandboxes” will help start-ups develop innovative AI systems in line with the new rules.

What does Europe do to support AI start-ups?

Legal certainty is important for start-ups, but regulation is only a small piece of the puzzle. We are making sure that European start-ups have access to the key ingredients of AI: data, computing power, algorithms and talent. Thanks to our data legislation and European data spaces, start-ups will have access to large volumes of industrial data to develop new AI applications. In addition, Europe has a world-leading public infrastructure of supercomputers. We are giving start-ups access to these supercomputers to help them train and develop their large AI models. And most importantly, we are boosting digital skills, fostering a new generation of pioneering AI researchers and entrepreneurs.

Does the AI has any disadvantages despite the fact that EU will introduce a law?

AI offers great opportunities for European citizens and businesses, but it can also be used in ways that run counter to our principles, values and rights.

While most uses of will pose little or no risk, certain AI systems can pose risks to safety and fundamental rights that need to be addressed. Examples are cyber security risks in the case of AI in critical infrastructure, risks of bias and discrimination in the case of AI used in recruitment processes, or health and safety risks in the case of AI used in medical devices. And some risks may still be unknown.

That is why in Europe we decided to act quickly and develop the world’s first comprehensive regulatory framework for AI, the AI Act. I believe we have struck the right balance to grasp the opportunities of AI, while managing its risks.

How could AI affect EU elections? Is there any risk of complications or even any danger the results to be affected?

It is true that we are seeing an increase in illegal content and disinformation, sometimes generated by AI (deepfakes), being spread across certain platforms in the EU. But you can be sure that we will make full use of our toolbox to protect our citizens and our democracies.

In Europe, we now have the DSA (Digital Services Act) and we will make sure that platforms play by the rules. In addition, the AI Act will contain strict requirements for generative AI systems to mark artificially generated content.

What are your next steps? A new mandate as an EU Commissioner or maybe an even higher ranking role?

As we say: there’s no time but the right time! And for the time being, I am fully focused on my mission as Commissioner, and with many files spanning from digital, space, industry and defence, there is plenty to do until the very last day of this mandate.