Relates to the use of automated lending decision-making tools by banks for the purposes of making lending decisions; allows loan applicants to consent to or opt out of such use.
This bill would add a new section to New York’s banking law regulating the use of automated lending decision-making tools by covered state-regulated lenders. It defines such tools broadly to include software using algorithms, computational models, or artificial intelligence that materially affects lending decisions, while excluding ordinary office software and internal administrative tools. The bill requires covered entities to conduct annual impact assessments, and to do so before any material change to a tool that could affect outcomes. Those assessments must evaluate the tool’s objectives, design, training data, accuracy, fairness, bias, discrimination, cybersecurity, privacy, public health and safety risks, misuse risks, and data practices, and the resulting summary must be posted on the lender’s website.
The bill also creates notice and consumer-protection requirements for loan applicants. Lenders using these tools must inform applicants in advance that an automated tool will be used, identify the criteria and data sources involved, explain data retention practices, and, if a loan is denied, provide the reason to the extent practicable. Applicants denied based on incorrect personal information would have 30 days to correct the information and appeal. The superintendent of financial services would gain authority to inspect records, investigate violations, require reports, and order a covered entity to stop using a tool if an assessment finds discriminatory or biased outcomes.
In addition, the bill would require covered entities to retain impact assessments for seven years and provide them to the department upon request. If a tool is found to produce discriminatory or biased outcomes, the entity must report that finding within 30 days, and the department must direct the entity to cease using the tool and any information produced by it. The bill also allows the superintendent to impose additional reporting obligations on entities that have deployed tools producing discriminatory or biased outcomes. The act would take effect 90 days after becoming law.
The overall sentiment reflected in the committee votes appears generally favorable, with the bill advancing through the Assembly Banks Committee and the Assembly Codes Committee by clear margins. That said, the recorded opposition suggests some concern about the scope of regulation, compliance burdens, or the practical effects of restricting automated underwriting tools. Because there are no transcripts provided, the specific arguments for or against the bill are not available, but the vote pattern indicates support for stronger oversight of AI in lending alongside some resistance.
The main point of contention is likely the balance between consumer protection and lender flexibility. Supporters appear to favor transparency, bias testing, and stronger safeguards against discriminatory lending, while opponents may be concerned about operational costs, disclosure requirements, and the possibility that the superintendent could effectively halt use of a tool after a bias finding. The bill is aimed at state-regulated banks and licensed lenders, and it expressly excludes certain federally chartered institutions from the definition of covered entity.
The bill would amend the Banking Law by adding a new section 103-a governing automated lending decision-making tools used by state-regulated banking organizations, foreign banking corporations licensed in New York, interstate branches authorized in New York, and licensed lenders. It would impose new compliance, disclosure, recordkeeping, and reporting duties on those entities, while giving the superintendent of financial services expanded oversight and enforcement authority. The measure would not apply to national banks, federal savings banks, federal savings and loan associations, federal credit unions, federal trust companies, or foreign banking corporations organized under U.S. law, and it would take effect 90 days after enactment.
The bill appears to have a generally favorable reception in committee, as shown by repeated favorable votes advancing it through the Assembly Banks and Codes Committees. The vote margins suggest meaningful support for regulating AI-driven lending decisions, likely driven by concerns about fairness, transparency, and discrimination. At the same time, the presence of several negative votes indicates that some members were not fully comfortable with the breadth of the requirements or the enforcement powers granted to the superintendent.
The central controversy is whether automated lending tools should be subject to extensive state oversight, including mandatory impact assessments, public reporting, applicant notice, and the possibility of being shut down after a bias finding. Supporters are likely focused on preventing discriminatory lending, improving transparency, and protecting consumers from opaque AI-driven decisions. Opponents are likely concerned about regulatory burden, the feasibility of detailed disclosures, the risk of overbroad interference with underwriting practices, and potential conflicts with federal banking regulation or preemption issues.