Poland advances AI oversight law to align with EU Artificial Intelligence Act

A committee of the Polish parliament has adopted a draft bill designed to align Poland’s national framework on artificial intelligence with the Artificial Intelligence Act.

The Committee on Digitalization, Innovation and Modern Technologies approved the proposal after introducing 20 amendments to the text.

One of the bill’s central elements is the creation of a new national oversight authority, the Commission for the Development and Security of Artificial Intelligence (KRiBSI).

According to the proposal, the body would supervise compliance of AI systems with national and EU rules. Its powers would include inspecting companies, verifying whether AI systems meet regulatory requirements, imposing penalties, and ordering withdrawal of non-compliant systems from the market.

The draft also introduces authorisation procedures for certain high-risk AI systems, including systems used in education, critical infrastructure, employment, and migration management.

Another component concerns regulatory sandboxes. The proposed authority would establish controlled testing environments where companies could develop and trial AI systems before broader deployment.

The commission would include representatives from multiple Polish regulatory bodies, including authorities responsible for competition, telecommunications, financial supervision, and broadcasting.

The bill also defines institutional responsibilities linked to EU implementation procedures. Poland’s digital affairs ministry would coordinate funding for AI research and development using national and EU resources, while conformity assessment notifications would fall under ministry responsibility and accreditation would remain with the Polish Centre for Accreditation.