In a historic milestone for the country's digital transformation, the bill for the implementation of the European Artificial Intelligence Act (AI Act) has entered the processing stage in the competent parliamentary committee of the Hellenic Parliament. This development is not merely a formal incorporation of a European directive; it is the foundation of a national framework that will define how technology interacts with citizens, the labor market, and public administration for decades to come. As artificial intelligence ceases to be a science fiction scenario and becomes an integral part of daily life, the need for clear "rules of the road" on the digital highway has become imperative.
The Architecture of Risk: What is Prohibited and What is Permitted
The bill follows the core philosophy of the AI Act, which is based on categorizing AI systems according to the level of risk they pose to fundamental rights and citizen safety. At the top of the pyramid are "unacceptable risks," which lead to a total ban on specific applications. These include social scoring systems—similar to those used in authoritarian regimes—as well as the use of real-time biometric identification in public spaces for law enforcement purposes, with very few and strictly defined exceptions related to national security or the prevention of terrorist acts.
Particular emphasis is placed on "high-risk" systems, which include algorithms used in healthcare, education, recruitment, and the administration of justice. For these cases, the bill provides for strict transparency requirements, data logging, and, most importantly, human oversight. The goal is clear: technology must never function as a "black box" that makes decisions about people's lives without the possibility of review and accountability.
The National Supervisory Authority and the State's Role
One of the most critical aspects of the bill is the designation of the National Supervisory Authority, which will undertake the heavy task of monitoring the implementation of the Regulation in Greece. The debate in Parliament focuses on whether this responsibility will be assigned to an existing independent authority, such as the Hellenic Data Protection Authority (HDPA), or whether a new body with specialized expertise will be created. This choice will determine the effectiveness of audits and the speed of response to rapid technological developments.
Furthermore, the bill introduces the concept of "regulatory sandboxes." These are controlled environments where companies, especially startups, can develop and test innovative AI systems under the supervision of the authorities before they hit the market. This measure aims to prevent innovation from being stifled under the weight of bureaucracy, allowing Greece to remain competitive on the global technology map.
Transparency and Deepfakes: The Battle Against Misinformation
In an era where AI-generated content can blur the lines between reality and fabricated truth, the bill imposes strict labeling rules. Any content generated by artificial intelligence (such as deepfake text, audio, or video) must bear a clear indication that it is the product of an algorithm. This is vital for protecting the democratic process and preventing the manipulation of public opinion, especially during election periods.
At the same time, rules are established for General Purpose AI models, such as ChatGPT. The creators of these models will be required to provide detailed summaries of the content used to train their systems, respecting the intellectual property rights of creators—an issue that has sparked intense controversy worldwide.
Conclusion: Towards Human-Centric Technology
The AI Act bill in Parliament is not the end of the road, but the starting point of a new era. The challenge for the Greek state is twofold: on the one hand, to protect citizens from the potential arbitrariness of algorithms, and on the other, to create a friendly environment for investment and research. The success of this endeavor will depend on staffing the supervisory authorities with the right personnel and continuous consultation with civil society and the academic community. Artificial intelligence is a tool; whether it will be a tool for progress or oppression depends on the legal and ethical framework we set today.
