On December 27th, the Armenian government approved the signing of the Council of Europe’s framework convention on “Artificial Intelligence and Human Rights, Democracy and the Rule of Law.” This crucial step marks a significant advancement in protecting human rights and developing a relevant legal framework in the field of artificial intelligence (AI) application.
While this initiative is of paramount importance, several risks and potential pitfalls may arise during both the convention’s implementation and the development of the associated legal framework. One key risk is inadequate legal and practical enforcement. Signing the convention does not guarantee its effective implementation, especially given existing weaknesses in domestic legal mechanisms. This could manifest as incomplete transposition or implementation of the convention’s provisions due to an ambiguous legislative base, weak oversight and sanction mechanisms potentially allowing for AI system abuses, and inconsistencies between local legislation and the convention’s requirements.
Another challenge lies in limited technological and institutional capacities. Armenia’s constrained technological development and limited capacity of oversight institutions could hinder the achievement of the convention’s objectives. This includes the absence of robust monitoring and evaluation mechanisms, potentially allowing for AI system violations, and insufficient capacity of state bodies to effectively oversee these technologies.
Economic risks and the potential impact on the business environment also warrant attention. Strict implementation of the convention’s provisions could create economic difficulties, particularly for small and medium-sized businesses (SMEs) that may struggle to meet AI regulation requirements. This could translate to an increased burden on small businesses due to additional regulations surrounding AI technology adoption and a decrease in economic investment due to strict oversight and sanctions.
International cooperation could also face obstacles. While signing the convention obligates Armenia to adhere to international standards, inconsistencies between these requirements and the regional political or economic situation could impede cooperation. This could result in strained dialogue and cooperation with international partners due to potential violations or discrepancies and a loss of competitiveness with regional countries if neighboring states have less stringent AI regulations.
Ensuring the exclusion of discrimination and the protection of human rights remains a crucial concern. Despite the convention’s focus on preventing discrimination, its full practical implementation may prove difficult. There is a risk of overlooking discrimination based on race, religion, or other grounds during the development or application of AI systems, as well as risks of biometrical data protection and privacy breaches.
A lack of a context-specific approach also poses a challenge. The convention’s general requirements may overlook local conditions and specificities. Imposing a single, uniform approach to AI application that does not correspond to Armenia’s existing technological and social conditions is a risk, as is the formalistic implementation of the convention’s provisions without considering sector-specific characteristics.
To mitigate these risks, several recommendations are offered: clarifying and harmonizing the legal framework by developing local legislation that aligns with the convention’s requirements and addresses gaps in Armenian legislation, developing technological capacities through investments in AI system oversight and impact assessment, raising public awareness through educational and awareness campaigns on the proper application of AI systems, leveraging international expertise by working with international partners and applying best practices, and establishing an independent oversight body to register violations and enforce legal requirements.
This framework convention, open for signature by both member and non-member states of the Council of Europe, represents the first international legal instrument aimed at regulating AI systems and protecting fundamental human rights. The convention was opened for signature on September 5, 2024, during the informal conference of justice ministers of Council of Europe member states in Vilnius. The convention aims to prevent any form of discrimination in digitalization processes related to race, religion, gender, or other grounds. It establishes appropriate oversight and accountability mechanisms, including penalties for violations. To date, the convention has been signed by the United Kingdom, Norway, Iceland, Moldova, Georgia, Andorra, San Marino, as well as observer states to the Council of Europe, the USA, and non-member Israel. The convention will enter into force after at least five states, including three Council of Europe members, ratify it.
Armenia currently has incomplete legal regulations in the field of AI application. Signing the convention will allow for the formation of a comprehensive legal framework in the field of AI application, which will promote human rights protection and the preservation of democratic principles. Signing the convention will also allow Armenia to cooperate with international partners through the exchange of legal regulations and best practices. This will not only increase the level of accountability of AI systems in Armenia but also contribute to the development of technologies without endangering human rights and the rule of law. This important initiative is in line with the challenges of the modern world, ensuring a responsible and safe digital environment.
