The concept of Artificial Intelligence (AI) has its roots in the mid-20th century and developed as an interdisciplinary endeavor that combined logic, mathematics, philosophy, and later computer science. One of the founding figures of this field was Alan Turing, who in his famous 1950 paper “Computing Machinery and Intelligence”, raised the fundamental question: “Can a machine think?” He proposed an experimental method to assess artificial intelligence, which later became known as the “Turing Test”, which aimed to discover whether a machine could exhibit behavior indistinguishable from that of a human being. In this paper, Turing proposed an experimental method to assess the intelligence of machines, which later became known as the “Turing Test”, with the aim of determining whether a machine could exhibit behavior indistinguishable from that of a human being. In his analysis, Turing also refers to Professor Geoffrey Jefferson’s position, expressed in a 1949 speech, according to which a machine could not be considered equal to the human brain as long as it was not able to create, feel and experience emotions authentically, but only to simulate logical processes through its mechanisms.[1]
The term “artificial intelligence” was first articulated in 1956 by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon, during the Dartmouth Conference in the USA. The authors proposed that “any aspect of learning or any other characteristic of intelligence can, in principle, be described so precisely that a machine can be simulated to perform it”.[2]
Today, AI takes on a de facto role as an informal legal actor, directly influencing the legal relations of citizens, often without human intervention or a transparent process that allows for control or appeal. Automated decision-making can not only reduce the role of institutional actors, but also undermine the human right to a fair and just trial, specifically guaranteed by Article 6 of the European Convention on Human Rights and Fundamental Freedoms.
In the wake of this change in modern life, in 2017,[3] the European Council called for “urgent sensitivity to emerging trends,”[4] including “issues such as artificial intelligence…, while ensuring a high level of data protection, digital rights and ethical standards.” In its 2019 Conclusions on the Coordinated Plan for the development and use of artificial intelligence “Made in Europe,”[5] the Council further stressed the importance of fully respecting the rights of European citizens and called for a review of existing legislation so that it adapts to the new opportunities and challenges posed by AI. The European Council also called for a clear definition of AI applications that should be considered high-risk.[6]
Along the same lines, the European Parliament has played an important role in shaping the EU’s approach to AI, taking a number of legislative and strategic initiatives. In October 2020, it adopted three important resolutions addressing the ethical aspects of artificial intelligence,[7] legal (civil) liability[8] and copyright protection.[9] These were followed in 2021 by other resolutions focusing on the use of AI in the criminal field[10] as well as in education, culture and the audiovisual sector.[11]
The culmination of these efforts was the adoption of Regulation (EU) 2024/1689 on Artificial Intelligence (AI Act) [12], which represents the first comprehensive attempt at global level to regulate the development, placing on the market and use of artificial intelligence systems.
Article 1, EU Artificial Intelligence Act, specifically provides: “The purpose of this Regulation is to improve the functioning of the internal market and to foster the use of human-centric and trustworthy artificial intelligence (AI), while ensuring a high level of protection of health, safety, fundamental rights as guaranteed by the Charter of Rights, including democracy, the rule of law and the protection of the environment, against the harmful effects of AI systems in the European Union, and by supporting innovation[…]”.
In line with Regulation (EU) 2024/1689 (AI Act) and within the framework of the obligations arising from the SAA, the CAA, as the competent institution, has drafted the Draft Law “On Artificial Intelligence”. This draft law largely follows the structure and philosophy of the AI Act, aiming to create a harmonized legal framework for the development, placing on the market and use of artificial intelligence systems in the Republic of Albania.
The purpose of the draft law is to improve the functioning of the market
and the legal environment in the Republic of Albania, promoting the development and use of human-centered and trustworthy artificial intelligence systems, as well as guaranteeing a high level of protection of health, safety, fundamental rights, democracy and the rule of law.
The draft law applies to providers, users, importers and distributors of artificial intelligence systems, as well as to manufacturers that integrate AI systems in their products. The scope of application also extends to entities located outside the territory of the Republic of Albania, when the results of their systems are used in Albania.
On the other hand, the draft law does not apply to AI systems used exclusively for military, defense or national security purposes, to models developed solely for scientific research and development, as well as to purely personal and non-professional uses by individuals.
Article 3 defines “AI system”, which means a machine-based system that is designed to operate with varying levels of autonomy and that can exhibit adaptability after being put into operation, and that, for express or implied purposes, derives, from the input data it receives, the means to generate results such as predictions, content, recommendations or decisions that can affect physical or virtual environments.
The draft law prohibits a series of artificial intelligence practices that are considered incompatible with the protection of fundamental rights and human dignity. These include manipulative systems, exploitation of individual vulnerabilities, social scoring systems, profiling to predict criminal behavior, the creation of biometric databases through massive data collection, as well as certain forms of biometric identification and emotion recognition.
Categorisation of artificial intelligence systems
In line with the approach of the European Union’s AI Act, the Albanian draft law classifies artificial intelligence systems according to the level of risk they pose.
High-risk AI systems
A system is considered to be high-risk when it is used as a security component of a product or when the system itself is subject to conformity assessment procedures before being placed on the market. Specific requirements are foreseen for these systems, such as risk management, drafting technical documentation, recording events, ensuring human oversight, respecting security standards and conducting an impact assessment on fundamental rights. They must also be provided with a declaration of conformity and registered in the relevant databases.
The draft law establishes transparency obligations for certain AI systems, requiring natural persons to be informed when they interact directly with an artificial intelligence system, except in cases where this is self-evident from the context of use.
General Purpose AI Models
The draft law also regulates General Purpose AI Models, imposing obligations on their providers related to technical documentation, transparency, sharing information with other developers and compliance with copyright legislation.
Systems-risk models
Additional requirements are foreseen for general-purpose AI models with systemic risk, such as advanced testing, identification and minimization of systemic risks, reporting of serious incidents and ensuring a high level of cybersecurity.
Authorized representative
Providers of AI models located outside the Republic of Albania must appoint an authorized representative on Albanian territory, who will ensure the fulfillment of the obligations provided for by the law.
The concept of a regulatory “sandbox”
One of the novelties of the draft law is the creation of a regulatory “sandbox”, a controlled environment where developers and providers of artificial intelligence systems can test their technologies before placing them on the market. This mechanism aims to foster innovation, while ensuring compliance with legal requirements and the protection of fundamental rights.
In conclusion, the draft law “On Artificial Intelligence” constitutes an almost complete transposition of Regulation (EU) 2024/1689 (AI Act) and is presented as an act of a highly technical nature. Although it represents an important step towards harmonisation with European legislation, its practical implementation is expected to be challenging, as it requires institutional capacities and qualified specialists in various technical and legal fields to guarantee effective supervision and implementation of its provisions.
[1] Turing, A. M. Computing machinery and intelligence., 1950, https://courses.cs.umbc.edu/471/papers/turing.pdf
[2] McCarthy, J., Minsky, M.,Rochester, N., & Shannon, C. E. (1955). A proposal for the Dartmouth Summer Research Project on Artificial Intelligence
[3] European Commission, Proposal for a Regulation of the European Parliament and of the Council on establishing harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain Union legislative acts, COM(2021) 206 final, 21 April 2021.
[4] European Council, European Council Meeting (19 October 2017) – EUCO Conclusions 14/17, 2017, p. 8.
[5] Council of the European Union, Artificial Intelligence b) Conclusions on the Coordinated Plan on Artificial Intelligence – Adoption, 6177/19, 2019.
[6] European Council, Special meeting of the European Council (1 and 2 October 2020) – Conclusions, EUCO 13/20, 2020.
[7] European Parliament resolution of 20 October 2020 on a framework for ethical aspects of artificial intelligence, robotics and related technologies, 2020/2012(INL).
[8] European Parliament resolution of 20 October 2020 on a civil liability regime for artificial intelligence, 2020/2014(INL).
[9] European Parliament resolution of 20 October 2020 on intellectual property rights for the development of artificial intelligence technologies, 2020/2015(INI).
[10] European Parliament preliminary report, Artificial intelligence in criminal law and its use by police and judicial authorities in criminal matters, 2020/2016(INI).
[11] European Parliament preliminary report, Artificial intelligence in education, culture and the audiovisual sector, 2020/2017(INI). In this regard, the Commission has adopted the Digital Education Action Plan 2021-2027: Reshaping education and training for the digital age, which foresees the development of ethical guidelines for the use of AI and data in education – Commission Communication COM(2020).
[12] European Union. (2024). Artificial Intelligence Act.
