Requires entities that decline access to important services (e.g. healthcare and financing) on the basis of an AI system's decision to evaluate the system for accuracy & discrimination, to disclose the system's use, and to enable affected individuals to appeal the decision to a human being.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding legal statute with mandatory obligations, enforcement mechanisms, and specific compliance requirements for controllers and processors engaged in automated decision-making.
The document primarily addresses discrimination and fairness risks (1.1, 1.3), with strong focus on automated decision-making systems affecting critical services. It also covers transparency issues (7.4) through disclosure requirements and governance mechanisms to prevent unfair outcomes.
The document governs AI use across multiple critical service sectors including Finance and Insurance, Health Care and Social Assistance, Real Estate, Educational Services, and Public Administration. It applies to automated decision-making systems that affect access to financial services, housing, healthcare, education, employment, insurance, and basic necessities.
The document primarily covers the Deploy and Operate and Monitor stages, with requirements for disclosure at deployment and ongoing annual impact assessments. It also addresses Verify and Validate through testing requirements for accuracy, fairness, bias and discrimination, and touches on Build and Use Model through requirements to evaluate training data.
The document explicitly covers automated decision-making systems and processes but does not use specific AI terminology like 'AI models', 'AI systems', 'frontier AI', or 'general purpose AI'. It focuses on automated processing systems used for consequential decisions. No compute thresholds or model types are mentioned.
New York State Legislature
This is an Act to amend the general business law of New York, indicating it was proposed by the New York State Legislature.
The document grants enforcement authority to state, federal, or local government authorities who can request and review impact assessments.
Government authorities at state, federal, or local levels are empowered to monitor compliance by requesting and reviewing impact assessments. Additionally, external independent auditors or researchers are required to conduct assessments.
The document explicitly targets 'controllers' and 'processors' engaged in automated decision-making affecting critical services. These are entities that deploy AI systems for decision-making purposes.
6 subdomains (3 Good, 3 Minimal)