What must system providers do before marketing high-risk AI systems according to EU regulations?

Under the EU regulations, system providers must register high-risk AI systems in a public database before they can be marketed. This requirement is part of the regulatory framework designed to enhance transparency and accountability in the use of AI technologies. By maintaining a public registry, the EU aims to ensure that high-risk AI systems are subject to oversight and scrutiny, allowing for better monitoring of the systems' compliance with ethical standards and safety requirements.

The public database serves several purposes: it aids in tracking the deployment of high-risk AI applications, facilitates regulatory compliance checks, and enhances public trust by ensuring that these systems are openly documented and accessible. This registration process is a crucial step in the overall governance of AI, helping to address potential risks associated with high-risk systems before they reach the market and start impacting users and society.

In contrast, conducting a user survey, implementing advanced security measures, or developing a marketing strategy, while important aspects of the product lifecycle, do not meet the specific regulatory requirement aimed at ensuring safety and accountability for high-risk AI systems prior to their commercial release.