NYC Local Law 28 — AI Tenant Screening Restrictions

Overview

This regulation entry addresses the conceptual framework of New York City Local Law 28, focusing on Artificial Intelligence (AI) in tenant screening, as implied by the provided document ID RAI-US-NY-NYCL28-2023. It is crucial to note that, based on official New York City legislative records, Local Law 28 of 2023 is officially titled "A Local Law to amend the administrative code of the city of New York, in relation to creating a mental health coordinator to inform city employees about mental health support and services," and was passed by the Council on December 20, 2023, and returned unsigned by the Mayor on January 19, 2024, taking effect in 120 days. Therefore, the specific designation "NYC Local Law 28 — AI Tenant Screening Restrictions" as a formal, enacted law for 2023 does not align with official numbering. However, New York City has been actively engaged in regulating AI and automated decision-making systems across various sectors, including housing and employment, indicating a strong policy interest in this area. This entry synthesizes the likely components of such a regulation, drawing from existing and proposed NYC legislation concerning AI, fair housing, and tenant protections, to provide a comprehensive overview of what a law with this title and focus would entail within the New York City regulatory landscape.

The intent behind a regulation like NYC Local Law 28 on AI tenant screening would be to address growing concerns about algorithmic bias, lack of transparency, and potential discrimination in the use of automated tools by landlords and property management companies. The rapid adoption of AI in tenant screening processes, which analyze various data points like credit history, income, and background checks, has raised alarms among housing advocates and policymakers regarding fairness and equitable access to housing. Such a law would aim to establish clear guidelines for the deployment and oversight of these technologies, ensuring that human rights and civil liberties are protected throughout the rental application process. It would likely build upon the principles established in other groundbreaking NYC AI regulations, such as Local Law 144 of 2021 concerning Automated Employment Decision Tools, adapting them to the unique context of residential housing. The overarching goal would be to foster a more transparent and accountable housing market where technology serves to enhance, rather than undermine, fair housing practices.

Definitions

For the purpose of a regulation on AI tenant screening, several key terms would require precise definitions to ensure clarity and enforceability. An "Automated Decision Tool" (ADT) or "Artificial Intelligence System" would likely be defined broadly to encompass any computational process, including machine learning, statistical modeling, data analytics, or AI, that generates a score, classification, or recommendation used to substantially assist or replace discretionary decision-making in tenant screening. This definition would aim to capture a wide array of algorithmic tools used to evaluate prospective tenants, from those assessing financial stability to those analyzing background information. This mirrors the broad definition used in NYC's Local Law 144 for employment tools.

A "Landlord" or "Housing Provider" would refer to any owner, lessor, sublessor, assignor, or managing agent of residential real property, or any other person or entity acting on their behalf, who uses an ADT in making decisions related to renting, leasing, or renewing a lease for a dwelling unit. "Tenant Screening" would be defined as any process by which a landlord evaluates a prospective tenant's eligibility, suitability, or desirability for a rental unit, including but not limited to, reviewing credit reports, criminal background checks, eviction history, income verification, and other personal data. "Bias Audit" would refer to an impartial evaluation by an independent auditor to assess an ADT for disparate impact on individuals based on protected characteristics under local, state, and federal anti-discrimination laws, such as race, gender, religion, national origin, disability, and familial status. These definitions would form the bedrock for establishing the scope and obligations outlined within the regulation.

Governance and Institutional Framework

The governance and institutional framework for overseeing AI tenant screening restrictions in New York City would likely involve existing municipal bodies with mandates related to housing and human rights. The New York City Commission on Human Rights (NYCCHR) would be a primary enforcement agency, given its role in investigating and prosecuting discrimination in housing under the NYC Human Rights Law. The NYCCHR already enforces anti-discrimination provisions and has experience with complex investigations related to housing practices.

Additionally, the Department of Housing Preservation and Development (HPD) could play a role in providing guidance and resources to landlords regarding compliance, similar to its involvement in other housing-related regulations. The Office of Technology and Innovation (OTI), which has been involved in developing AI action plans for the city, could potentially contribute technical expertise and coordinate inter-agency efforts, especially regarding the technical aspects of AI system evaluation and auditing. The establishment of a dedicated office for algorithmic accountability, as proposed in the broader GUARD Act, could also centralize oversight for all city-wide AI initiatives, including those impacting housing. This multi-faceted approach would leverage existing expertise while potentially creating new structures to effectively manage the complexities of AI regulation in the housing sector.

Key Focus Areas

A central focus of NYC Local Law 28 on AI tenant screening would be to mandate transparency and disclosure. Landlords would be required to provide clear and conspicuous notice to prospective tenants when an automated decision tool is used in the screening process. This notice would detail the specific criteria the AI tool evaluates, the data sources it uses, and how the tool contributes to the final rental decision. The goal is to move away from "black box" decisions, where applicants are denied without understanding the underlying reasons. This requirement would empower tenants by giving them crucial information to understand and potentially challenge adverse decisions, aligning with broader consumer protection and data rights principles.

Another critical area would be the prevention of algorithmic bias and discrimination. The law would likely mandate independent bias audits of AI tenant screening tools to ensure they do not produce disparate impacts based on protected characteristics, such as race, national origin, gender, or disability. These audits would need to be conducted regularly by impartial third parties, with summaries of the results made publicly available. Furthermore, the regulation would likely establish a right for applicants to request an alternative screening process or a reasonable accommodation if they believe the AI tool has unfairly evaluated them. This would provide a crucial safeguard against discriminatory outcomes and ensure that individuals are not unfairly excluded from housing opportunities due to flawed or biased algorithms. The law would also likely impose restrictions on the types of data AI tools can consider, particularly sensitive personal information or factors that have historically led to discriminatory practices.

Implementation Framework

The implementation framework for NYC Local Law 28 on AI tenant screening would involve a multi-pronged approach to ensure compliance and effective oversight. Landlords and property management companies utilizing automated decision tools for tenant screening would be required to register these tools with a designated city agency, such as the New York City Commission on Human Rights or a newly established Office of Algorithmic Data Accountability. This registration process would likely involve submitting documentation detailing the AI tool's functionality, its data inputs, the methodology for its bias audits, and a summary of audit results. The aim is to create a public record of AI tools in use and facilitate regulatory monitoring.

Furthermore, the regulation would likely stipulate technical and operational requirements for AI tenant screening tools. This could include standards for data quality, model explainability, and regular performance evaluations to ensure the tools remain fair and accurate over time. Landlords would also be responsible for establishing clear internal policies and training programs for their staff on the ethical and compliant use of AI in screening. The law might also encourage or mandate the development of standardized templates for applicant notices and adverse action letters, ensuring consistency and clarity in communication. Enforcement guidance and rules would be promulgated by the relevant city agencies to provide detailed instructions on how to meet the legal obligations, including specific methodologies for conducting bias audits and reporting requirements.

Monitoring and Evaluation

Effective monitoring and evaluation would be crucial for assessing the impact and compliance of NYC Local Law 28. The designated oversight agency, likely the New York City Commission on Human Rights, would be tasked with collecting data on the use of AI tenant screening tools across the city. This data collection could include the number of applications processed by AI, the demographic breakdown of applicants and accepted tenants, the frequency and outcomes of bias audits, and the number of complaints received regarding AI-driven screening decisions. Regular reporting from landlords on their AI tool usage and audit results would be a cornerstone of this monitoring effort, enabling the city to identify trends, potential areas of concern, and systemic issues.

Beyond data collection, the monitoring framework would involve periodic reviews of the law's effectiveness in achieving its objectives of transparency, fairness, and non-discrimination. This could include public hearings, stakeholder consultations with tenant advocates and housing providers, and independent academic studies to evaluate the real-world impact of the regulation. The city might also establish a mechanism for reviewing and updating the technical standards and best practices for AI tenant screening tools as technology evolves. The goal of this continuous monitoring and evaluation process would be to ensure that the law remains relevant and effective in protecting tenants' rights in an increasingly AI-driven housing market, allowing for adaptive regulation in response to new challenges and technological advancements.

Penalties, Liability, and Appeals

To ensure compliance and provide recourse for aggrieved parties, NYC Local Law 28 would establish a clear framework for penalties, liability, and appeals. Landlords or housing providers found in violation of the regulation, such as failing to provide required disclosures, using unaudited AI tools, or engaging in discriminatory practices through AI, would be subject to civil penalties. These penalties would likely be structured to deter non-compliance, potentially including escalating fines for repeat offenses, similar to the penalty structure for Local Law 144 on employment AI, which ranges from $500 to $1,500 per violation. Each instance of non-compliance, such as a failure to provide notice to an individual applicant, could constitute a separate violation.

Regarding liability, the law would likely hold landlords directly responsible for the compliant use of AI tenant screening tools, even if those tools are developed or provided by third-party vendors. This would place the onus on housing providers to conduct due diligence and ensure that any AI systems they deploy adhere to the regulation's requirements. Tenants who believe they have been harmed by a violation of the law would have the right to file a complaint with the New York City Commission on Human Rights. The Commission would then investigate the complaint and, if a violation is found, could order remedies such as monetary damages, injunctive relief (e.g., requiring the landlord to cease using a non-compliant AI tool), or other appropriate actions. An appeals process would also be established, allowing both landlords and tenants to challenge decisions made by the enforcement agency through administrative review and, if necessary, judicial review in the appropriate courts.

Relationship to Other Instruments

NYC Local Law 28 on AI tenant screening would operate within a complex web of existing legal instruments at the local, state, and federal levels. It would build upon and complement the foundational protections offered by the New York City Human Rights Law, which broadly prohibits discrimination in housing based on numerous protected characteristics. The local law would specifically extend these anti-discrimination principles to the context of automated decision-making in tenant selection, addressing the unique challenges posed by algorithmic bias.

Furthermore, this regulation would interact with New York State laws governing landlord-tenant relations, such as the Housing Stability and Tenant Protection Act of 2019 (HSTPA), which introduced significant changes to tenant rights, including restrictions on eviction history use in screening. It would also need to align with federal statutes like the Fair Housing Act, which prohibits housing discrimination, and the Fair Credit Reporting Act (FCRA), which regulates the collection and use of consumer report information, including credit and background checks used in tenant screening. The Fair Chance for Housing Act (Int. No. 632-A), passed by the NYC Council in December 2023, which limits the use of criminal background checks in tenant screening, would be particularly relevant, as AI tools often incorporate such data. The AI tenant screening law would likely integrate with these existing frameworks, adding a layer of AI-specific requirements without superseding the broader protections already in place. It would also draw parallels from NYC Local Law 144 of 2021, which regulates Automated Employment Decision Tools, adapting its principles of bias auditing and transparency to the housing context.

International Alignment

While NYC Local Law 28 is a municipal regulation, its principles and objectives align with a growing international movement to regulate artificial intelligence, particularly concerning its impact on fundamental rights and equitable access to essential services. Jurisdictions worldwide, including the European Union with its proposed AI Act, are developing frameworks that emphasize transparency, accountability, and human oversight for high-risk AI systems. Tenant screening, given its potential to impact access to housing, would likely be categorized as a high-risk application of AI under many emerging international standards.

The focus on bias audits, disclosure requirements, and the right to explanation in NYC's conceptual AI tenant screening law reflects a global consensus on the need to mitigate algorithmic discrimination and ensure fairness. Many countries and international organizations, such as the OECD, have issued recommendations and guidelines for the responsible development and deployment of AI, underscoring principles like human-centricity, robustness, and accountability. Although a local law in New York City would not directly enforce international treaties or regulations, its design would implicitly draw from and contribute to these broader global conversations about ethical AI governance, particularly in sensitive sectors like housing. The city's proactive stance in regulating AI in housing could also serve as a model for other municipalities and national governments grappling with similar challenges.

Implementation Timeline

Compliance Checklist

Sources and References