A manifesto on AI in law was presented in Europe. The UNBA has joined the initiative

On June 24, as part of European Legal Sovereignty Day, which took place in Brussels, the Manifesto for a Referential of Human Guarantee in the Legal Use of AI was presented.

On behalf of the Ukrainian National Bar Association, the declaration of accession to the Manifesto was signed online by the President of the UNBA, BCU Lidiya Izovitova.

The Manifesto, prepared by the National Bar Council of France (CNB) in collaboration with Legal Data Space and the Delegation of French Bar Associations, is based on two structural requirements that form the common foundation of any legal practice integrating AI:

1) human responsibility. Any significant legal decision must remain within the sphere of responsibility of a legal professional. Artificial intelligence can provide analysis, research, or decision preparation, but it cannot replace either human judgment or human responsibility;

2) effective oversight. Human oversight cannot be purely theoretical. It must actually provide the ability to understand, verify, modify, or reject the result generated by an automated system.

The authors of the Manifesto derive both of these requirements from the provisions of Regulation (EU) 2024/1689 of the European Parliament and of the Council of June 13, 2024, establishing harmonized rules on artificial intelligence (the Artificial Intelligence Regulation, AI Act).

The Manifesto is also consistent with the position of the Council of Bars and Law Societies of Europe (CCBE), as set forth in the Position Paper dated October 8, 2021, regarding the proposed regulation on harmonized rules in the field of artificial intelligence. The CCBE consistently advocates for the preservation of human oversight and the professional responsibility of advocates as indispensable principles for the use of artificial intelligence in legal practice.

Currently, legal analysis, document drafting, legal research, and decision-support functions are already partially automated, and this change is structural and long-term. At the same time, human oversight remains fragile: at times purely formal, at other times excessive. In the absence of a common framework that specifically defines the role of humans in AI-enhanced legal processes, human oversight risks losing its significance as a safeguard.

An advocate’s professional interest lies in defining the limits of permissible AI use while ensuring the preservation of attorney-client privilege and personal data, as well as recognizing personal responsibility for an outcome that cannot be entirely dependent on technology.

The Ukrainian National Bar Association has taken the initiative to participate in the development, testing, and implementation of technical safeguards for human oversight a practical Human-in-the-Loop framework for legal artificial intelligence. This framework is intended to define common, specific, and operational conditions for human oversight in AI systems used in legal and judicial activities.

The Ukrainian advocacy association has been working systematically in this area for several years. Within the UNBA, there is a working group on the legal regulation of artificial intelligence, which analyzes legal issues related to the development and use of AI and prepares proposals for regulatory frameworks that ensure adequate protection of human rights during the development, implementation, and use of artificial intelligence systems.