Sponsors (5)
Official title: Establish protections for individual rights with respect to computational algorithms, and for other purposes.
Introduced December 2, 2025 by Edward John Markey · Last progress December 2, 2025
The bill strengthens individual privacy, transparency, anti‑discrimination protections, and enforcement for consequential automated systems, but does so at the cost of substantial compliance and litigation burdens, regulatory uncertainty, and risks to trade secrets that could disadvantage small developers and slow deployment of some beneficial services.
People subject to high‑stakes automated decisions — especially racial and ethnic minorities, people with disabilities, low‑income communities, immigrants, and other marginalized groups — gain stronger protections against discriminatory outcomes because the bill treats consequential actions as covered and bans algorithms that cause disparate impact.
Individuals (including people with disabilities, low‑income people, and immigrants) get clearer notice, plain‑language disclosures, accessible formats, contact information, and explicit rights to human alternatives and appeals for consequential automated decisions, improving transparency and recourse.
Consumers and the public benefit from mandatory independent pre‑deployment and annual evaluations, documentation, and independent audits that increase auditability and reduce the risk of harmful or unsafe algorithmic behavior before wide deployment.
Developers, deployers, and ultimately consumers — especially small businesses and startups — will face substantial new compliance costs for audits, documentation, accessible disclosures, contract duties, and penalty exposure, which could raise prices or limit offerings.
Smaller developers and startups will be disproportionately disadvantaged because independent audits, documentation retention, and contractual obligations are resource‑intensive compared with large firms that can absorb costs.
Broad, flexible definitions (e.g., 'consequential action', 'covered algorithm') and lengthy agency rulemaking discretion create regulatory uncertainty that could chill innovation, produce uneven compliance, and delay clear rules for businesses and individuals.
Based on analysis of 12 sections of legislative text.
Requires audits, impact assessments, disclosures, and FTC enforcement for commercial algorithms used in high‑stakes consequential actions to prevent discrimination and harms.
Creates a federal rulebook for commercial “covered algorithms” used in high‑stakes areas (hiring, housing, credit, health care, criminal justice, elections, benefits, etc.). It bans uses that cause disparate impacts or discrimination, requires pre‑deployment and annual impact assessments (with independent auditors for likely harms), mandates public disclosures and recordkeeping, gives the FTC enforcement authority (including new penalty amounts and state parens patriae suits), and funds agency hiring and a new federal algorithm‑auditing occupational series at OPM. Developers and deployers must share documentation, consult affected communities, prevent and mitigate harms identified by evaluations, and keep reports and public summaries for years; the FTC gets rulemaking and enforcement powers, including expanded jurisdiction over some previously exempt entities.