Overview
Severity: MEDIUM | Affected: European Union | Category: policy
The European Parliament has officially passed the 'AI Model Liability Act' (AMLA), a landmark piece of legislation that establishes a framework for accountability concerning damages caused by autonomous AI systems. The act introduces a tiered system of responsibility, placing the primary onus on developers of 'high-risk' AI systems, as defined in the original AI Act. For the first time, the AMLA creates a legal presumption of causality, making it easier for victims of AI-caused harm (e.g., from autonomous vehicles or medical diagnostic AI) to seek compensation. The legislation requires developers to maintain detailed, auditable logs of their models' decision-making processes and training data provenance. Critics argue the act could stifle innovation, while proponents hail it as a critical step in ensuring trustworthy AI and protecting consumers in an increasingly automated world. The act is set to take effect in January 2027.