Amends state finance law to require state units purchasing a product containing algorithmic decision systems to purchase only those adhering to responsible AI standards. Outlines such standards to include harm avoidance, transparency, fairness, and risk evaluation.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding legislative act that amends New York state finance law and executive law with mandatory procurement requirements for state units purchasing algorithmic decision systems. The bill uses mandatory language ('shall purchase') and establishes enforceable legal obligations.
The document has minimal to good coverage of approximately 5-6 subdomains, with primary focus on unfair discrimination (1.1), privacy compromise (2.1), lack of transparency (7.4), and lack of robustness (7.3). Coverage is concentrated in discrimination/fairness, transparency requirements, and AI system reliability domains.
This bill governs AI procurement across all sectors of New York state government operations. The primary sector governed is Public Administration (excluding National Security), as the bill applies to state units including governmental agencies, political subdivisions, and public benefit corporations. The governance extends to any sector where state government operates or procures AI systems.
The document primarily focuses on the procurement and deployment stages of AI systems by state units. It establishes requirements for purchasing algorithmic decision systems that adhere to responsible AI standards, with emphasis on transparency, fairness, harm avoidance, and risk evaluation. The document does not explicitly address data collection, model building, or ongoing operational monitoring stages.
The document explicitly defines and covers 'algorithmic decision systems' which encompasses AI systems broadly. It does not specifically mention frontier AI, general purpose AI, foundation models, generative AI, predictive AI, open-weight models, or compute thresholds. The focus is on algorithmic decision systems that impact individuals, regardless of specific AI type or architecture.
New York State Assembly; M. of A. SOLAGES
The bill was introduced by Member of Assembly SOLAGES and referred to the Committee on Governmental Operations, indicating it was proposed by the New York State Legislature.
commissioner of taxation and finance
The commissioner of taxation and finance is explicitly designated with regulatory authority to implement and enforce the provisions of this subdivision.
commissioner of taxation and finance
While not explicitly stated as a monitoring role, the commissioner of taxation and finance's regulatory authority implicitly includes monitoring compliance with the procurement standards established by the bill.
state units; governmental agencies; political subdivisions; public benefit corporations of the state
The bill explicitly targets state units purchasing algorithmic decision systems. State units are defined as 'the state and any governmental agency or political subdivision or public benefit corporation of the state.'
6 subdomains (2 Good, 4 Minimal)