They recommend a risk-based approach that recognizes the need for greater oversight in AI systems with legal or significant societal impacts.



Source link

author-sign