On 19 February 2025, the Arkansas Digital Responsibility, Safety, and Trust Bill (SB 258) was introduced to the Senate, including design requirements. The Bill mandates responsible development and deployment of artificial intelligence (AI) systems, particularly those classified as high-risk. High-risk AI systems are defined as those that make or substantially influence decisions producing legal or similarly significant effects concerning consumers, such as financial services, housing, healthcare, employment, and criminal justice. Developers of high-risk AI systems must take reasonable measures to prevent algorithmic discrimination and provide documentation on risks, data sources, and mitigation strategies. They must share relevant information with deployers and the Attorney General upon request and ensure transparency through public summaries of their AI systems and risk management practices. Any known risks of algorithmic discrimination must be disclosed promptly, with updates required within 90 days of significant modifications.
Original source