CLEAR Act

Bill Number: S. 3813
Origin Chamber: Senate
Congress: 119th Congress, Session 2
Policy Area: Commerce
Status: Introduced
Latest Action: 2026-02-10: Read twice and referred to the Committee on the Judiciary.
Last Updated: 2026-02-27T19:56:01Z

AI-Generated Summary

Purpose

The CLEAR Act aims to promote transparency in the development of generative artificial intelligence (AI) by requiring developers to disclose the use of copyrighted materials in training their models. This helps copyright owners track and protect their works while fostering ethical AI practices.

Key Provisions

Definitions:
Artificial intelligence: An automated system mimicking human-like tasks or thinking.
Copyrighted work: Creative content protected by U.S. copyright law (Title 17) and either registered or scheduled for registration.
Generative AI model: Software using AI to create outputs like text, images, audio, or video.
Training dataset: A collection of materials (e.g., text, images) used to teach the AI model.
Register: The head of the U.S. Copyright Office.

Notice Requirement: Anyone using a training dataset containing copyrighted works to train or release a generative AI model must submit a notice to the Register. The notice includes:
A detailed summary of each copyrighted work in the dataset.
A web link (URL) to the dataset if it's publicly available online.

Timing for Submission:
For models first used or released after the law's effective date: At least 30 days before commercial use (including internal business use) or any release.
For existing models (used or released before the effective date): Within 30 days after the Register issues implementing rules.

Regulations: The Register must create rules within 180 days of the law's enactment, specifying the notice's format, details, and submission process.

Enforcement:
Copyright owners can sue in federal court if a notice is not submitted.
Possible court remedies include:
Civil penalties of at least $5,000 per violation, paid to the Copyright Office to cover its costs (capped at $2.5 million total per year per violator).
Court orders (injunctions) to stop using the copyrighted work until notice is filed.
Reimbursement of the owner's legal fees.
The annual penalty cap does not limit access to injunctions or fee awards.

Public Database: The Register must maintain an online, publicly accessible database of all submitted notices.

Effective Date: The law takes effect 180 days after being signed into law.

Significant Changes to Existing Law

This bill introduces a new mandatory disclosure requirement under U.S. copyright law (Title 17), which previously did not specifically address AI training data. It adds a pre-use reporting obligation for generative AI developers and creates a private right of action (ability to sue) for copyright owners focused on transparency violations, separate from traditional infringement claims. It also directs penalties to fund the Copyright Office, a novel funding mechanism not previously tied to AI.

Potential Impacts

Government Agencies: The U.S. Copyright Office will face increased administrative duties in processing notices, issuing regulations, and maintaining a database, potentially offset by penalty revenues. This could enhance the Office's role in overseeing emerging technologies.
Citizens and Businesses: AI developers (e.g., tech companies) may incur compliance costs for reviewing datasets and filing notices, possibly slowing innovation or raising development expenses. Copyright owners (e.g., artists, writers) gain tools to monitor and enforce rights, potentially leading to more resolved disputes without full lawsuits. The public benefits from greater transparency about AI training sources, which could inform ethical AI use.
International Relations: Minimal direct impact, though it may influence global AI standards by setting a U.S. precedent for copyright disclosure in AI, affecting multinational tech firms or cross-border data flows.

Main Stakeholders Affected

AI Developers and Companies: Bear the primary compliance burden, including dataset audits and notice submissions.
Copyright Owners: Authors, musicians, filmmakers, and other creators who can now more easily detect and challenge unauthorized use of their works in AI training.
U.S. Copyright Office: Responsible for implementation, enforcement support, and public database management.
General Public and Consumers: Indirectly affected through increased AI transparency and potential shifts in how AI-generated content is created and regulated.

Notable Legal, Constitutional, or Political Implications

Legal: Strengthens copyright enforcement in the AI era by bridging gaps in existing law, potentially increasing litigation over data transparency without altering core infringement rules. The penalty structure encourages compliance while limiting excessive fines, but it could lead to more court cases testing what constitutes a "sufficiently detailed summary."
Constitutional: No apparent conflicts with free speech or due process, as the law targets disclosure rather than content creation; however, it may raise questions about the scope of copyright in transformative AI uses.
Political: Bipartisan sponsorship (by Sens. Schiff and Curtis) signals broad support for balancing AI innovation with intellectual property protection. It could spark debates on regulating Big Tech versus supporting creators, influencing future AI policy amid growing concerns over data ethics and job displacement in creative industries.

This summary was generated by AI and may contain inaccuracies. Refer to the official source document for the authoritative text.

Sponsor

Sen. Schiff, Adam B. [D-CA]

Cosponsors (1)

Sen. Curtis, John R. [R-UT]

Recent Actions

2026-02-10: Read twice and referred to the Committee on the Judiciary.
2026-02-10: Introduced in Senate