The European AI Act and the debate between standards and innovation

Table of contents

The wonderful video illustrating the extraordinary capabilities of Google’s innovative artificial intelligence called Gemini is a fake. By Google’s own admission, the video that went viral in a very short time and shows how Gemini would be able to react to moving images was artfully constructed, in short, a clumsy attempt that resulted in what we could define as a striking example of artificial-artificial intelligence. The artifice that creates the illusion of something that is promoted as artificial intelligence, has very little intelligence and too much artificial intelligence, at least in this case. However, the matter has been accomplished: the video has been made, the mystery has been revealed, the faults have been admitted, the end of the story, perhaps there remains some reputational damage for Google that most of us will forget within a few days. Because, however, we are interested in this fact, because it shows once again that clear rules are needed so that innovation can express itself at its best and without undesirable consequences. It goes without saying that the issue of video is minimal, cited here in pure exemplary function, if compared to the potential of artificial intelligence, with its development capabilities and variety of applications. That’s why we need rules. And here two doors open: the one that opens onto the room where the goodness and effectiveness of the rules are discussed, and the one that instead opens onto the room where the now traditional theme is discussed, which sees the factions of those who want innovation at any cost and who argue that the excess of rules slows down the innovation process and those who instead support the need for clear and defined rules, otherwise innovation risks get out of control.

AI Act

Now the issue is fundamental when it comes to artificial intelligence, even after the approval by the European Parliament and the Member States of the so-called AI Act, in fact the first regulatory framework in the world that regulates the uses and applications of AI. The project has been talked about for months, we wrote about it in June , but only now after marathons to find agreements on the various points has the final text been reached. The chronicle of the facts is reported by many (Euronews for example ), a chronicle that emphasizes the efforts made to reach the agreement, the limitations that the AI Act places on the use of artificial intelligence in relation to the reading of emotions and above all on the use of biometric data, with the sole exceptions of terrorist events, serious crimes, search for victims and missing persons as a result of criminal acts. Limits are also placed on the use of so-called predictive policing and the use of AI for security, although compared to the initial text, the limitations have also been softened as a result of specific positions, such as that of France, which is developing AI-based solutions for security and surveillance to be used during the Olympic and Paralympic Games in Paris scheduled for next year. The chronicles continue by reporting the joy of Thierry Breton, European Commissioner for the Internal Market, who commented on the event as a historic moment, which in fact it is since the EU is the first governmental organization in the world to approve a document that regulates the use of artificial intelligence. The AI Act also provides for very heavy penalties that can reach fines of up to 35 million euros or equal to 7% of the turnover of companies and organizations that violate the legislation and also precedes, as is natural, a continuous update following the development of new derivations, possibilities, technological applications based on artificial intelligence. The approval of the AI Act is also an act that fits into the furrow that the EU has already traced with the Digital Service Act, the Digital Market Act, the Data Act that we wrote about here also according to the impact they have on the work of startups .

Rules and innovation

Let’s close the door for the moment on the room where the details of the regulations are being discussed, since there is now an approved text, let’s focus on the room where the pro- and anti-regulation factions confront each other. Here the issue is very open and will always be: on the one hand there are those who say that it is essential to regulate, that Europe plays a role as a world leader on this issue, the most striking example is that of the GDPR on the protection of privacy that has also inspired other governments on the planet, on the other hand there are those who fear that regulation risks becoming a brake on innovation and makes Europe lose further ground compared to to the rest of the world, especially the USA and China, where there is no regulation and therefore it is easier to innovate and develop new technologies and applications. It is the search for balance between maker and shaper (we wrote about it here ) that lies at the heart of this issue. The debate is very open, on the one hand there are those who argue that it is true that in other countries since there are no stringent rules, innovation has the opportunity to run faster, on the other hand there are those who support the importance of fully implementing democratic tools capable of protecting individuals, tools that in other countries do not exist with the related consequences. A good compromise, or rather a good model of approach is the one proposed by Ian Hogarth, who in the meantime has been appointed to lead the British taskforce on AI, as we wrote here , it is a model that is inspired by that of CERN in Geneva, therefore the creation of a defined space where experiments can be conducted without regulatory limits, A space, therefore, where innovation can express itself to its full potential, and then verify that when developed it is actually compatible with society and everyday life before making it available on the market.

Startups in Europe

Finally, there is a further element that can be added to this framework: the creation of a European regulatory framework for startups. It is an element that would fit perfectly into the approach that the EU is having and that would give impetus to innovation made by entrepreneurs because if it is true that regulations can be an element of brake on innovation, it is also, and above all, true that having uncoordinated and non-homogeneous regulations as it is today for startups in Europe, represents a multiplication of this brake. Therefore, also in view of the upcoming European elections, a proposal to harmonize the way in which European startups develop, raise funds, and go to market, could represent an almost unique opportunity to demonstrate how the regulatory aspect can effectively become a driving force for innovation. (Photo by Marius Oprea on Unsplash)

ALL RIGHTS RESERVED ©

    Subscribe to the newsletter