Poland's Sejm Passes AI Systems Law, Sends It to the President

Poland's Sejm (the lower house of parliament) completed work on the law on artificial intelligence systems on July 3, transposing the EU's AI Act into Polish law. Lawmakers adopted 24 of 25 amendments proposed by the Senate, rejecting one by a vote of 405 to 13 with 17 abstentions. The bill now goes to the president for signature.

A long legislative path

The bill reached the Sejm on April 9, after the Council of Ministers (Rada Ministrów) approved it on March 31. The lower house passed the law on June 11 by a vote of 421 in favor, 3 against and 18 abstentions. The Senate submitted its amendments on June 25, approving the bill unanimously with 79 votes in favor. Rapporteur Grzegorz Napieralski reported that the Sejm's committee on digitalization, innovation and modern technologies reviewed the Senate's amendments on July 2 and recommended adopting 24 of the 25.

A new oversight body

The centerpiece of the law is the creation of the Komisja Rozwoju i Bezpieczeństwa Sztucznej Inteligencji (Commission for the Development and Security of Artificial Intelligence, KRiBSI), which will become the main body overseeing Poland's AI market. The commission will conduct administrative proceedings, accept and review complaints from citizens, carry out inspections of high-risk systems, and issue orders to withdraw non-compliant systems from the market.

Penalties and regulatory sandboxes

The law provides for steep financial penalties, up to 35 million euros or 7 percent of a company's global annual turnover in the most serious cases of violations, and up to 15 million euros or 3 percent of turnover in lesser cases. Obstructing inspections or failing to comply with the commission's decisions also carries criminal liability, including the possibility of imprisonment. At the same time, the law introduces regulatory sandboxes, controlled environments in which companies can test AI systems under state supervision without risking immediate sanctions.

Deputy Digitalization Minister Dariusz Standerski stressed that the regulatory sandboxes are meant to give businesses room to test solutions in a friendly legal environment rather than operating under legal uncertainty. Deputy Prime Minister Krzysztof Gawkowski said the law is meant to ensure AI is no longer seen as a set of incomprehensible algorithms, but becomes an area governed by clear, transparent rules.

What it means for businesses and citizens

For companies deploying AI systems, especially those classified as high-risk, this means preparing for KRiBSI inspections and facing real financial risk in the event of violations. The regulatory sandboxes, however, could give startups and larger companies a way to legally test new AI applications before they reach the broader market.

Citizens gain the right to file complaints directly with the commission and access to public information about its decisions and the sanctions it imposes. Oversight of high-risk systems is also reinforced by the president of the Personal Data Protection Office (Urząd Ochrony Danych Osobowych, UODO), who receives new powers under the law. Once signed by the president, Poland will become one of the first EU countries with a complete national law implementing the AI Act, which could influence the pace at which other member states finalize their own regulations.

Sources: End of parliamentary work on the artificial intelligence law. Time for the president to act (cyberdefence24.pl), Sejm adopts artificial intelligence law. AI commission and regulatory sandboxes for businesses (rp.pl), AI Act. Commission for the Development and Security of Artificial Intelligence (kancelaria-szip.pl)