The European legislation on artificial intelligence approved a few days ago represents, for the first time in the world, an attempt by an institution to balance the benefits of technological progress with the necessary protection of the lives of citizens at large: in short, this legislation embodies the concern not to let AI become a monster capable of eating "pieces" of human existences.

With an emphasis on respect for European values and the importance of putting humans at the center of artificial intelligence development, the AI Act introduces clear prohibitions and transparency obligations to ensure ethical and responsible use of this emerging technology.

The main rules established by the AI Act for the development and use of artificial intelligence in Europe include:

1. European Approach to Artificial Intelligence.: The AI Act builds on the European approach to artificial intelligence, which promotes excellence and trust as pillars to enhance research and industry while ensuring security and protection of fundamental rights.
2. Prohibition of Facial Recognition: The regulation prohibits real-time facial and biometric recognition in public areas without exception.
3. Transparency Obligations for Generative Models.: All generative artificial intelligence models, including those not at systemic risk, are subject to transparency requirements regarding their creation and technical characteristics.
4. Risk Classification.: The AI Act identifies risk profiles in the use of artificial intelligence, distinguishing between high-risk and low-risk systems.
5. Clear Responsibilities for Companies: The regulation outlines clear responsibilities and requirements for companies based on the level of risk of artificial intelligence systems, directly influencing technological innovation in the European market.
6. Governance and Effective Application: The AI Act stipulates that member states play a key role in the governance and enforcement of artificial intelligence legislation, ensuring respect for fundamental rights and security.

The main differences between the AI Act and previous laws on the use of artificial intelligence in Europe can be summarized as follows:

1. Horizontal and Risk-Based Approach.: The AI Act takes a horizontal, non-sectoral approach and is based on the concept of risk. This uniform regulation will apply directly in all member states without the need for national transposing laws.
2. Facial Recognition Prohibited: One of the innovations introduced by the AI Act is a complete ban on real-time facial and biometric recognition in public areas, with no exceptions; to give an example, it will be impossible for administrations or police organizations to "read" the identities of passersby through biometric facial detection for the simple purpose of recording citizens' movements.
3. Regulation of Generative AI Models: The AI Act regulates generative artificial intelligence models, not just the end products (applications), recognizing the importance of regulating the technology itself to prevent risk.
4. Transparency Requirements for Generic Models: The regulation introduces transparency requirements for generic artificial intelligence models, such as self-assessments, systemic risk mitigation, major incident reporting and cybersecurity requirements.
5. Broader Application: The AI Act extends not only to suppliers introducing systems to the European market but also to users, importers, distributors and all parties involved in the use of such systems, including those outside the EU.

This is a first, big step that will make it possible to avoid the jungle, the Wild West that could be the impact of AI experienced totally unaware...