Congress mandates AI audits and civil rights protections for algorithmic decisions

What it does

This bill creates a federal civil rights framework for artificial intelligence algorithms used in high-stakes decisions affecting employment, housing, credit, education, voting, law enforcement, and other consequential actions. It requires developers and deployers of AI systems to conduct pre-deployment audits and annual impact assessments by independent auditors, disclose their practices publicly, avoid discrimination and disparate impact on protected groups, and provide individuals with rights to appeal algorithmic decisions and opt for human review. The FTC gains authority to enforce these requirements and promulgate detailed rules.

Why we flagged it

The bill's core mechanism is a civil rights framework imposing pre-deployment audits, transparency, and anti-discrimination duties on AI developers and deployers. It is functionally a regulatory regime for algorithmic accountability, not a deregulation or carve-out.

What the text implies

• The definition of 'consequential action' is extremely broad (Section 2(4)) and includes voting, elections, and election administration—potentially subjecting election-related algorithms to FTC audit and disclosure requirements, raising federalism and state election administration questions.
• The 'disparate impact' standard (Section 2(10)) presumes algorithms are not separable for analysis and shifts burden to developers/deployers to prove otherwise by preponderance of evidence—a significant evidentiary burden that may chill algorithmic innovation in high-stakes domains.
• Independent auditor requirements create a new market for compliance services but also a potential bottleneck: the bill does not specify auditor qualifications, conflicts-of-interest standards are strict (Section 2(12)(B)), and no safe harbor exists for good-faith compliance efforts.
• The 10-year retention requirement for evaluations, assessments, and contracts (Sections 102(e)(2), 202(b)(2)) creates substantial data storage and privacy obligations, particularly for developers handling sensitive personal data across multiple deployers.
• Section 203(a) requires the FTC to promulgate regulations on opt-out and human alternatives within 2 years, but the bill does not specify what 'reasonable' opt-out mechanisms are or whether deployers can charge for human review—leaving significant implementation uncertainty.

Section numbers refer to the bill text the analysis read — linked under Primary records below.

Who it affects

Ordinary people gain enforceable rights to know how AI systems affect them, to appeal algorithmic decisions, and to opt for human review in high-stakes contexts (employment, housing, credit, voting, law enforcement). The bill creates accountability mechanisms and transparency requirements that address real harms from algorithmic bias and discrimination. Compliance costs to industry do not outweigh the public benefit of civil rights protections.

Who stands to gain

• independent auditors and compliance consultants
• legal and compliance service providers
• technology firms specializing in algorithmic transparency and bias detection

Named in the bill

Federal Trade Commission (FTC), developers (AI algorithm creators), deployers (entities using AI algorithms), independent auditors, individuals (natural persons in the US), protected characteristic groups (race, color, ethnicity, national origin, religion, sex, disability,

Where it stands

29 cosponsors: 29 Democrats.

• Dec 2, 2025 — Introduced · Congress.gov: “Introduced in House”
• Dec 2, 2025 — Referred to House Committee on Oversight and Government Reform and House Committee on Energy and Commerce · Congress.gov: “Referred to the Committee on Energy and Commerce, and in addition to the Committee on Oversight and…”

Dates and quoted wording are Congress.gov's action record; the timeline shows status changes, not every procedural step.

Money around this bill

2 lobbying clients named this bill on 5 disclosure filings across 2 quarters, Dec 2025 to Jun 2026. Those filings disclosed $5,605,837 in lobbying spend. A filing names 39 bills on average, so that figure is what each filing reported, not a share belonging to this bill.

More lobbying clients named this bill than 41% of bills with at least one filing.

Yvette Clarke, the sponsor, reported $626,107 in PAC receipts in the 2026 cycle.

• American Council of Life Insurers — $4,085,837 on 3 filings
• AFL-CIO — $1,520,000 on 2 filings

Lobbying Disclosure Act filings through Jul 20, 2026. A filing shows who paid to lobby on a bill it names, not what changed.