Artificial intelligence; Responsible Deployment of AI Systems Act; AI Council; AI Regulatory Sandbox Program; Artificial Intelligence Workforce Development Program; effective date.
HB1916 creates the “Responsible Deployment of AI Systems Act” and establishes a statewide framework for regulating artificial intelligence systems used by public entities and private organizations in Oklahoma. The bill defines AI systems and sorts them into four risk categories: unacceptable, high risk, limited risk, and minimal risk. Unacceptable-risk systems would be prohibited, while high-risk systems would face the most stringent requirements, including pre-deployment assessments, independent audits, ongoing monitoring, and qualified human oversight. Limited-risk systems would have transparency obligations, and minimal-risk systems would be subject to basic data protection standards.
The bill requires deployers of AI systems to classify systems before use, assess bias and safety risks, document mitigation strategies, maintain inventories and audit records, and notify affected individuals when high-risk AI influences decisions. It also requires accessible feedback channels, annual performance reporting, and public disclosure of noncompliance incidents. In addition, the bill creates the Artificial Intelligence Council to oversee implementation, issue guidance, administer an AI regulatory sandbox for testing innovative systems, and run a workforce development program focused on AI training for small businesses, educational institutions, and underserved communities.
If enacted, HB1916 would add new provisions to Title 25 of the Oklahoma Statutes and impose direct compliance duties on public agencies and private deployers of AI systems operating in the state. It would require organizations to build internal governance, recordkeeping, audit, and human-review processes for higher-risk AI uses, and it would authorize penalties, suspension of noncompliant systems, and mandatory external oversight for violations. The bill would also create a new state AI Council with rulemaking authority and responsibilities related to oversight, sandbox approvals, workforce grants, and annual reporting.
The available record shows no committee transcript or vote history, so there is no documented floor or committee debate to gauge support or opposition. Based on the bill text, the measure is framed as a comprehensive consumer-protection and accountability proposal that also promotes innovation and workforce development, suggesting an intent to balance regulation with economic opportunity. Because the bill had only been referred to Rules as of the last action, its political reception in the legislative process is not yet clear from the provided materials.
The most likely points of contention are the breadth of the regulatory scheme and the compliance burden it places on both public and private deployers of AI. Potentially controversial provisions include the prohibition on “unacceptable risk” systems, mandatory independent audits, human oversight requirements, public disclosure of noncompliance, and the AI Council’s authority to issue rules and penalties. Supporters would likely emphasize civil rights, privacy, safety, and transparency, while critics may focus on costs, administrative complexity, and whether the definitions and risk categories are too expansive or vague.