Bill > HB3567

This bill establishes the Meaningful Human Review of Artificial Intelligence Act, which sets strict guidelines for how state agencies can use automated decision-making systems (AI systems that make or support decisions without full human intervention). The bill requires that these systems cannot be used for functions related to public assistance benefits, individual rights, or decisions that significantly impact citizens' welfare without continuous meaningful human review. This means a trained human with the authority to modify or override the AI's decisions must be directly involved in the process. State agencies must conduct comprehensive impact assessments before implementing such systems, which must evaluate the system's objectives, accuracy, potential biases, cybersecurity risks, and privacy concerns. These assessments must specifically test for discriminatory outcomes based on characteristics like race, gender, disability, and other protected attributes. If an assessment finds discriminatory results, the agency must immediately stop using the system. Additionally, the bill protects state employees by ensuring that AI systems cannot replace human workers, reduce their hours, or alter their employment terms. Impact assessments must be submitted to state leadership, published online, and updated at least every two years or whenever the system undergoes significant changes, providing transparency and ongoing oversight of AI technologies used in government services.

Military Affairs and Security

https://www.ilga.gov/legislation/BillStatus.asp?DocNum=3567&GAID=18&DocTypeID=HB&SessionID=114&GA=104