Policy Brief Library

The current administration's foundational AI executive order, establishing a federal policy of minimal regulatory burden on AI development, creating a DOJ AI Litigation Task Force to challenge state AI laws, and directing federal agencies to develop a preemptive national legislative framework. The primary document defining the federal government's current stance on AI regulation — and the most significant obstacle to meaningful public protection.

The Biden administration's comprehensive AI safety executive order, directing federal agencies to adopt safety standards, protect civil rights, and establish AI risk management frameworks. Though revoked in January 2025, it remains the most comprehensive federal attempt to protect the public from AI harm and a model for future legislation.

Required all federal agencies to implement AI governance frameworks consistent with the NIST AI Risk Management Framework, mandating protections for civil rights and safety in government AI deployments by December 2024. An important benchmark for evaluating federal agency compliance — and a document watchdog groups have found is frequently ignored.

The primary voluntary federal framework for AI risk management in the United States, establishing standards for trustworthiness, accountability, transparency, and bias mitigation. Though non-binding, it has become the operational standard for federal agencies. Essential background for understanding what federal standards exist — and why voluntary frameworks are insufficient.

An extension of the core AI RMF specifically addressing the unique and heightened risks posed by generative AI — including hallucination, deepfakes, data poisoning, and harmful content generation. The first federal framework document to grapple specifically with the risks of the AI systems now reaching hundreds of millions of users.

Legislation making AI developers and deployers legally responsible for ensuring their systems do not violate civil rights laws — including the Civil Rights Act of 1964, the Fair Housing Act, the Voting Rights Act, and the ADA. Addresses the critical gap between existing civil rights law and AI-driven discrimination in hiring, housing, lending, and government services.

The first bipartisan federal law directly targeting AI-generated harm, criminalizing the knowing publication of non-consensual intimate imagery including AI-generated deepfakes. Though narrow in scope, it represents the first congressional acknowledgment that AI-specific legislation is necessary — and a model for expanding protections to other forms of AI-enabled harm.

A comprehensive bipartisan Senate proposal addressing AI safety testing, transparency, and accountability — the closest Congress has come to comprehensive AI regulation. Though not passed, it is expected to be reintroduced and remains the primary legislative template for federal AI accountability legislation.

Congressional Democrats' formal position on worker protections in AI regulation, demanding collective bargaining rights, transparency in AI workplace deployment, prohibitions on AI discrimination, and the right of workers to refuse AI-generated directives that violate their rights. A key document for the labor dimension of AI accountability advocacy.

The most comprehensive state AI law in the United States, requiring developers and deployers of high-risk AI systems to use reasonable care to protect consumers from algorithmic discrimination. Establishes disclosure obligations, impact assessment requirements, and a right to appeal AI-driven decisions. A model for what federal legislation should achieve and proof that comprehensive AI accountability is legislatively achievable.

Makes it a civil rights violation under Illinois law to use AI for employment decisions — including recruitment, hiring, promotion, and discharge — without notice to employees, or in a manner that discriminates based on protected characteristics. The first state law to establish a private right of action for AI-driven employment discrimination.

Texas's comprehensive and accountable law, applying to developers and deployers of AI doing business in the state. Prohibits AI systems designed for restricted purposes including encouragement of self-harm, unlawful discrimination, and CSAM generation. Demonstrates that AI accountability legislation has bipartisan support across ideological lines.

A policy article documenting the wave of nearly 700 AI-related state bills introduced in 2024 and arguing that the proposed federal moratorium on state AI laws would strip consumers, workers, and vulnerable communities of their primary legal protections in the absence of federal legislation. Essential reading for understanding the stakes of the federal preemption battle.

The world's first comprehensive legally binding AI regulation, establishing a risk-based framework that classifies AI systems into four tiers — from unacceptable risk (prohibited) to minimal risk. Bans social scoring, real-time biometric surveillance, and manipulation of vulnerable groups. With penalties up to €35 million or 7% of global turnover, it is the most consequential AI regulation in existence and the global standard against which U.S. inertia must be measured.

The most widely endorsed international AI governance framework, with 40+ adherent countries. Establishes the core principles — inclusive growth, human-centered values, transparency, robustness, and accountability — that serve as the international reference point for national legislation and provide the U.S. with a framework it has endorsed but not enacted into law.

The international agreement signed at the inaugural AI Safety Summit at Bletchley Park, establishing shared recognition that frontier AI poses serious risks requiring urgent international cooperation. The United States is a signatory — creating a baseline international commitment that no current domestic policy actively undermines.

The ACLU's comprehensive policy brief documenting how AI-driven decisions in hiring, housing, welfare, credit, and government services violate civil rights — and making the case for the AI Civil Rights Act as the legislative solution. Essential reading for understanding why existing civil rights law is insufficient to address algorithmic discrimination.

A policy framework proposing a multi-agency approach to AI regulation by sector-specific agencies with specific rules for each sector, stronger interagency coordination, more targeted consumer and civil rights protections than a single horizontal law could achieve. A practical roadmap for how federal regulation could be structured.

Congressional testimony from Consumer Reports documenting the consumer harm from deepfake technology and calling for stronger legislative protections — offering specific policy recommendations that consumer advocates can take to lawmakers. Articulates the full range of AI-generated deception targeting American consumers.