EU AI Act

EU AI Act: Social Impact and Regulations on AI

How Europe is redrawing the lines of digital sovereignty and safety

Entered into force on 1 August 2024, the AI Act is the first-ever legal framework on AI, not only addressing the risks of AI but also positioning Europe as a global regulatory leader. The AI Act will be fully applicable on 2 August 2026 ensuring that Europeans can trust what AI has to offer. As information and decisions are increasingly shaped by AI systems, the risks linked to their use are now at the centre of public and scholarly debate. In offering contributions to solve societal challenges, most AI systems pose limited to no danger; certain AI systems, though, create risks that we must address to avoid undesirable outcomes.

The Act in depth

According to the European Commission, the AI Act introduces a risk-based regulatory framework that classifies systems into four categories to safeguard public safety and trust. At the highest level are unacceptable risks, including AI applications that threaten fundamental rights - such as manipulative practices - which are therefore prohibited. AI systems classified as high risk, including those used in critical infrastructure, education, and law enforcement, may be deployed only if they meet stringent standards for data integrity, transparency, and human supervision. A third-tier targets transparency risks, requiring systems like chatbots and deepfakes to be clearly identified so users can make informed decisions. Finally, the minimal or no risk category cover the vast majority of AI tools currently in use, such as spam filters and video games, which remain unregulated in order to foster innovation.

These bans, which took effect in February 2025 and will be fully implemented in 2026, are essential if we want to coordinate a tool that could become larger than us. Sometimes we take its validity for granted, ignoring that AI tools make it easy for anyone to create fake images and news that could contribute to misinformation. Those with ill intentions can mass-produce and disseminate propaganda on social media.

A Threat to Democracy

To assure the protection of democracy and information, we must regulate the power of AI instruments. Individuals should be able to access reliable knowledge and understand the global context, while preventing the misuse of AI to distort information or limit these rights. ​​The goal of the EU AI Act of promoting reliable AI is one of the strongest moves that the EU has made to protect free information in these times. The Act forbids practices that endanger the personal privacy of individuals, such as social scoring. By this, we mean systems that use artificial intelligence and data analysis to assign people or organisations a numerical score based on their behaviours, characteristics, or interactions. These scores have the potential to influence access to services, benefits, and opportunities, thereby impacting an individual's treatment in society.

By dividing people into categories of “worthiness,” equality, fairness, and human fundamental rights are undermined. Furthermore, another protection provided by the Act concerns the use of AI systems in the administration of justice. Within democratic processes, artificial intelligence applications are classified as ‘high risk,’ meaning that they are subject to obligations and must be reviewed under human supervision in order to prevent discrimination that could influence elections or legal decisions. The boundary must be set where innovation undermines democratic principles: no creation should restrict our ability to make decisions or access information.

Setting a “human-centric” standard

Time will tell how effective this Act will be. In my opinion, it could represent an essential revolution in the regulation of cutting-edge technologies. By establishing a balanced relationship between what AI can be used for and what must remain reliant on human capacities, we stand to benefit from both change and innovation. The EU AI Act seems to be a perfect solution for continuing to take advantage of a tool that is now part of everyday life, while still monitoring its safety. As we navigate this new era, the priority remains clear: to maintain a human-centric system, and the Act provides the legal certainty needed to support this objective. By requiring appropriate human oversight for high-risk applications and prohibiting practices that pose a threat to fundamental rights and safety, the EU reinforces the principle that digital progress must not come at the expense of human dignity. Control over our decisions ought to rest with ourselves: this is where AI will never surpass us, in the capacity to direct our will as we choose.