Int 0926-2024

Be it enacted by the Council as follows: Section 1. Subdivision a of section 20-u of the New York city charter, as added by a local law for the year 2025 amending the New York city charter and the administrative code of the city of New York, relating to establishing an office of algorithmic accountability, as proposed in introduction number 199, is amended by adding new definitions of "artificial intelligence" and "public-impacting artificial intelligence" in alphabetical order to read as follows: Artificial intelligence. The term "artificial intelligence" has the same meaning as set forth in subsection (3) of section 9401 of title 15 of the United States code. Public-impacting artificial intelligence. The term "public-impacting artificial intelligence" means any artificial intelligence that could reasonably be expected to materially impact the rights, liberties, benefits, safety or interests of the public, including the public's access to available city services and resources for which they may be eligible. � 2. Section 20-u of the New York city charter, as added by a local law for the year 2025 amending the New York city charter and the administrative code of the city of New York, relating to establishing an office of algorithmic accountability, as proposed in introduction number 199, is amended to add a new subdivision e to read as follows: e. The office shall promulgate rules establishing basic compliance standards that all agencies must meet in developing, procuring, deploying, and using public-impacting artificial intelligence. Such rules shall include reporting requirements for agencies to document compliance with such standards. Such standards shall be reviewed no less than every other year and updated by rule as necessary. Such standards shall include, but need not be limited to: 1. Procedures for ensuring fairness, transparency, and accountability in public-impacting artificial intelligence decision-making processes by evaluating the fairness of any determinations, and, to the extent technically feasible, an evaluation of the process and data inputs used to produce them, including but not limited to standards for evaluating the design, application, and outcomes of public-impacting artificial intelligence to guard against bias; 2. Procedures for identifying, assessing, and mitigating risks associated with the use of public-impacting artificial intelligence, including but not limited to data protection risks; 3. Procedures for regular monitoring and evaluation of public-impacting artificial intelligence; 4. Procedures for protecting individual privacy and civil liberties; and 5. Procedures for verifying that an agency complies with such standards when such agency uses any public-impacting artificial intelligence that such agency did not develop, does not maintain, or does not control. � 3. This local law takes effect on the same date as a local law for the year 2025 amending the New York city charter and the administrative code of the city of New York, relating to establishing an office of algorithmic accountability, as proposed in introduction number 199, takes effect. CoJM/IB LS #12427, 13440 10/21/25 11:02 pm 2 2