The EU AI liability framework aims to address which of the following characteristics of AI?

The EU AI liability framework primarily aims to address the characteristics of opacity and complexity inherent in AI systems. This focus arises from the understanding that AI technologies often operate in ways that are not easily interpretable by users or even creators, leading to significant challenges in accountability and liability when these systems cause harm or fail to perform as intended.

Opacity refers to the difficulty in understanding how an AI system arrived at a certain decision or action due to its reliance on complex algorithms, often described as "black boxes." This complexity can hinder stakeholders from identifying who is responsible for outcomes caused by AI systems, which is why the liability framework seeks to clarify these issues.

By emphasizing this aspect, the framework aims to ensure that affected individuals can seek redress and that developers and operators of AI systems maintain a level of transparency and accountability that aligns with the broader goals of safety and ethical use of technology in society. Hence, focusing on opacity and complexity is crucial for establishing clear liability standards in the context of emerging AI technologies.