An independent evaluation, often recommended or required by an AI Accountability Framework, that assesses an AI system's actual or potential impact on individuals, society, or fundamental rights, ensuring compliance with ethical guidelines and regulatory requirements like the EU AI Act.

‍