Which obligation is specific to GPAI "with systematic risk" under the AI Act?

The obligation to assess and mitigate systemic risks is a critical component specific to General Purpose AI (GPAI) systems categorized under the AI Act with systematic risk. This is because GPAI technologies have the potential to impact individuals and society at large, given their widespread application and the inherent complexity associated with their deployment.

Systematic risks refer to risks that can affect entire systems or sectors rather than just individuals or organizations. By assessing these risks, developers and implementers of GPAI can identify potential adverse effects on users and society, enabling them to take proactive measures to control or reduce these risks. This obligation ensures that there is a structured approach to understanding and addressing any negative implications that might arise from the use of such AI systems.

In the context of the AI Act, the emphasis on risk assessment and mitigation aligns with broader goals of ensuring safety, protecting fundamental rights, and fostering trust in AI technologies. It goes beyond standard operational practices, as it specifically addresses the complex interplay between technology and societal frameworks, necessitating a higher degree of scrutiny and responsibility.

The other options, while valuable for business operations, do not specifically address the systemic risks associated with GPAI under the AI Act, making them irrelevant in this context.