EU Parliament's committees propose changes to AI Act

Przeczytasz w 2 minuty

On April 21, POLITICO published a leaked draft report on the AI Act. The report has been prepared jointly by the IMCO and LIBE committees of the European Parliament, with the support of MEPs Brando Benefei, and Dragos Tudorache, the chairman of the specially appointed Committee on artificial intelligence - AIDA. The document proposes changes to the AI Act aiming at increasing the fundamental and civil rights protection, imposing greater obligations on AI users, and introducing changes for the public sector AI management systems.

An important change is the requirement to register high-risk AI systems with the public administration in publicly available databases before their first launch. People whose sensitive data are used in high-risk AI systems should be informed about it. Databases should be transparent, accessible and readable to all users. In addition, it has been proposed that public administration has to publish the results of AI systems impact assessments on fundamental human rights. The right to privacy and protection of personal data should be guaranteed throughout the whole life cycle of AI systems, in accordance with the principle of data minimization and anonymization. People affected by the use of AI in the areas of health, safety or fundamental rights will have the right to complain and claim compensation.

On 18 February 2022, the Moje Państwo Foundation sent a letter to MEP Brando Benefei, proposing the amendments to the AI Act related to the public sector and fundamental rights. Our suggestions indicated, inter alia, issues related to strengthening the legal situation of persons harmed by AI, the creation of publicly available databases containing information about artificial intelligence systems used by the public administration, the obligation to inform citizens about the use of AI systems and its possible consequences in the public sector. 

The proposals for amendments to the AI act sent by the Foundation were mostly taken into account in the aforementioned IMCO – LIBE draft report, although the committee changes concern only high-risk AI systems. An important positive development is the proposed ability of the citizen to assert her/his rights in connection with the use of AI systems and their resulting negative consequences.

Details of the Foundation's position can be found here.

There are still legal gaps in the AI Act, namely in the context of:
  • definition of artificial intelligence;
  • registration of all AI systems used by public administrations – not only high-risk systems;
  • regulating the prohibition of using obtained biometric data in public spaces;
  • informing people about the use of AI systems in the public sector before making a decision;
  • the right of people not to be subject to AI systems violating the AI Act.

The proposed changes to the AI ACT can be tabled as amendments by MEPs until May 18.