AZ HB4098

Requires artificial intelligence businesses using "frontier models" (AI trained with over 10^26 computational operations or costing more than $100 million) to conduct comprehensive risk assessments for high-risk AI systems before use or sale

Mandates transparency reports be submitted to the Arizona Attorney General and publicly posted, detailing the AI system's purposes, risk assessment results, and bias mitigation measures

Requires quarterly audits to identify and correct discrimination, harmful outputs, or unsafe behavior, plus special protections including parental monitoring tools and content filtering for AI systems accessible to children

Attorney General must issue certification before an AI business can enter commerce and may impose civil penalties up to $50,000 per violation, with each day constituting a separate violation

Defines "high-risk AI systems" as those impacting children, healthcare, employment, public safety, or education, or likely to cause discrimination, bias, or significant privacy violations