NY, US · AI law tracker
A9654 — NY, US
A9654 is an AI governance legislation from NY, currently committee. New York's A9654, the AI Civil Rights Act, has been introduced and is now under review in the Assembly Committee on Science And Technology. AIGI tracks 1 primary-source update on this bill; the most recent was published on 2026-01-21.
Status & timeline
- Regulatory stage
- committee
- Bill status
- In Assembly Committee
- Authority / governing body
- New York State Senate
- Chamber
- Assembly
- Document type
- legislation
Next deadline: No fixed deadline for enactment; the bill is currently in committee.
Subscriber only
Full obligation matrix
| Actor | Obligation | Deadline | Source |
|---|---|---|---|
| deployer | Prohibited from using covered algorithms that discriminate based on protected characteristics. | — | — |
| deployer | Prohibited from using covered algorithms that cause a disparate impact. | — | — |
| developer | Prohibited from offering or licensing covered algorithms that discriminate based on protected characteristics. | — | — |
| developer | Prohibited from offering or licensing covered algorithms that cause a disparate impact. | — | — |
| developer | Complete independently audited pre-deployment evaluations to identify, evaluate, and mitigate potential biased use or discriminatory outcomes. | — | — |
| deployer | Complete independently audited pre-deployment evaluations to identify, evaluate, and mitigate potential biased use or discriminatory outcomes. | — | — |
| developer | Complete independently audited post-deployment impact assessments to identify, evaluate, and mitigate potential biased use or discriminatory outcomes. | — | — |
| deployer | Complete independently audited post-deployment impact assessments to identify, evaluate, and mitigate potential biased use or discriminatory outcomes. | — | — |
| developer | Mitigate any harms identified by pre-deployment evaluations and impact assessments. | — | — |
| deployer | Mitigate any harms identified by pre-deployment evaluations and impact assessments. | — | — |
| developer | Ensure that any covered algorithm performs reasonably well and is consistent with its publicly-advertised purpose. | — | — |
| deployer | Ensure that any covered algorithm performs reasonably well and is consistent with its publicly-advertised purpose. | — | — |
| deployer | Increase transparency around the use of covered algorithms in consequential decisions. | — | — |
| deployer | Provide individuals a right to appeal an algorithmic decision to a human decision-maker. | — | — |
Subscriber only
Enforcement risk score
25
/ 100
Announced regulation; enforcement footprint still forming.
Subscriber only
Role-based compliance checklist
- general_counsel Review the full text of A9654 once available to understand the precise definitions of 'covered algorithms', 'consequential decisions', and 'protected characteristics'. (todo)
- compliance_officer Inventory all AI/ML systems currently used in or planned for consequential decision-making processes, categorizing them by impact area (e.g., HR, finance, customer service). (todo)
- cto Assess technical capabilities for conducting independent pre-deployment evaluations and post-deployment impact assessments for potential bias and discriminatory outcomes. (todo)
- data_scientist Develop methodologies for identifying, evaluating, and mitigating algorithmic bias and disparate impact across various AI systems. (todo)
- hr_director Evaluate hiring and employee management systems using AI for compliance with anti-discrimination and transparency requirements, specifically the human appeal right. (todo)
- product_manager Ensure that AI products or features perform consistently with their publicly advertised purpose and do not introduce unintended discrimination. (todo)
- risk_manager Integrate algorithmic bias and discrimination risks into the overall enterprise risk management framework. (todo)
- board_director Oversee the organization's strategy for addressing AI civil rights, compliance, and risk mitigation. (todo)
Subscriber only
Vendor impact assessment
- Vendor risk class
- critical
- Procurement categories
- hr_tech, customer_service_ai, fraud_detection, financial_decisions, healthcare_ai
Vendors developing or deploying AI/ML solutions in consequential decision-making areas must demonstrate robust compliance with anti-discrimination, audit, impact assessment, harm mitigation, and human oversight requirements. New York-based businesses should prioritize vendors who can provide detailed evidence of such compliance measures.
Sample vendor questions
- How do you ensure your AI systems do not discriminate based on protected characteristics or cause disparate impact?
- Do you provide independent pre-deployment evaluations and post-deployment impact assessments for your covered algorithms, and can you share results?
- What processes are in place to identify and mitigate harms from biased use or discriminatory outcomes in your AI systems?
- How do you ensure your algorithms perform reasonably well and consistently with their advertised purpose?
- Do your AI systems support a human review and appeal process for consequential decisions?
- What transparency features are built into your AI systems regarding their use in consequential decisions?
Intelligence briefs (1)
legislation In Assembly Committee 1/21/2026
New York AI Civil Rights Act Bill Advances in Assembly Committee
New York's A9654, the AI Civil Rights Act, has been introduced and is now under review in the Assembly Committee on Science And Technology.
This bill signals an expanding legislative focus on AI accountability, particularly regarding discrimination, impact assessments, and human oversight of consequential algorithmic decisions.
Deadline: No fixed deadline for enactment; the bill is currently in committee.
Primary source →Frequently asked questions
- What is A9654?
- New York's proposed Artificial Intelligence Civil Rights Act (A9654) has advanced to the Assembly Committee on Science And Technology. The bill seeks to protect individual rights from computational algorithms, prohibiting discrimination by AI in consequential decisions affecting employment, banking, and healthcare [1]. It mandates independently audited pre-deployment evaluations and post-deployment impact assessments to mitigate bias, alongside requirements for transparency and a right to human appeal for algorithmic decisions [2]. Primary source →
- Why does A9654 matter?
- This bill signals an expanding legislative focus on AI accountability, particularly regarding discrimination, impact assessments, and human oversight of consequential algorithmic decisions. Primary source →
- Who does A9654 affect?
- This proposed legislation directly impacts developers and deployers of AI systems operating within New York State, particularly those whose algorithms are involved in consequential decisions. This includes applications in employment, banking, healthcare, the criminal justice system, public accommodations, and government services, requiring a focus on non-discrimination and algorithmic fairness. Primary source →
- What are the key dates for A9654?
- No fixed deadline for enactment; the bill is currently in committee. Primary source →
- What is the current status of A9654?
- As of the last published update, A9654 is at the "committee" stage, with bill status "In Assembly Committee". Primary source →
- Where can I find the primary source for A9654?
- The primary source for the most recent update is at https://legislation.nysenate.gov/bills/2025/A9654. AIGI publishes the full citation chain plus every approved brief on this bill. Primary source →
Related
Stay informed