WHAT THIS BILL REGULATES · 1 REQUIREMENT TYPE
How Is This Bill Enforced
Verbatim statutory text on the left; plain-language analysis and a per-section checklist on the right. Numbered markers cross-link to the matching checklist row.
This Act may be cited as the "Generative AI Copyright Disclosure Act of 2024".
Establishes the short title of the Act as the Generative AI Copyright Disclosure Act of 2024. No compliance obligations are created by this section.
(a)(1) 1 A person who creates a training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6), or alters a training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6) (including by making an update to, refining, or retraining the dataset) in a significant manner, that is used in building a generative AI systemGenerative AI systemThe term "generative AI system" means a software product or service that— (A) substantially incorporates one or more generative AI models; and (B) is designed for use by consumers.Sec. 2(d)(4) shall submit to the RegisterRegisterThe term "Register" means the Register of Copyrights.Sec. 2(d)(5) a notice that contains— (A) a sufficiently detailed summary of any copyrighted worksCopyrighted workThe term "copyrighted work" means a work protected in the United States under a law relating to copyrights.Sec. 2(d)(2) used— (i) in the training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6) (in the case that the person creates the dataset); or (ii) to alter the training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6) (in the case that the person alters the training data in a significant manner); and (B) the URL for such dataset (in the case of a training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6) that is publicly available on the internet at the time the notice is submitted).
(a)(2) 1 The notice required by paragraph (1) shall be submitted— (A) not later than 30 days before the generative AI systemGenerative AI systemThe term "generative AI system" means a software product or service that— (A) substantially incorporates one or more generative AI models; and (B) is designed for use by consumers.Sec. 2(d)(4) with respect to which the training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6) is used is made available to consumers, in the case that the generative AI systemGenerative AI systemThe term "generative AI system" means a software product or service that— (A) substantially incorporates one or more generative AI models; and (B) is designed for use by consumers.Sec. 2(d)(4) is first made available to consumers after the date on which this Act takes effect; and (B) not later than 30 days after the date on which this Act takes effect, in the case that the generative AI systemGenerative AI systemThe term "generative AI system" means a software product or service that— (A) substantially incorporates one or more generative AI models; and (B) is designed for use by consumers.Sec. 2(d)(4) with respect to which the training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6) was used was made available to consumers before the effective date of this Act.
This subsection imposes the bill's central obligation: any person who creates or significantly alters a training dataset used to build a generative AI system must file a notice with the Register of Copyrights. The notice must include a sufficiently detailed summary of copyrighted works used and, where the dataset is publicly available online, its URL. The filing deadline depends on whether the generative AI system is launched before or after the Act's effective date — 30 days before the system is made available to consumers for new systems, or 30 days after the effective date for existing systems.
The obligation is notable for its breadth: it applies to any person, not just the developer of the AI system, and covers both initial dataset creation and significant alterations such as updates, refinements, and retraining.
(b)(1) Any person described under paragraph (1) of subsection (a) that fails to comply with a requirement under such subsection shall be assessed a civil penalty in an amount not less than $5,000.
(b)(2) 2 Not later than 180 days after the date on which this Act takes effect, the RegisterRegisterThe term "Register" means the Register of Copyrights.Sec. 2(d)(5) shall issue regulations to implement the requirement under paragraph (1).
This subsection establishes the enforcement mechanism: a civil penalty of not less than $5,000 for any person who fails to comply with the notice requirement. No maximum penalty amount is specified. The Register of Copyrights is directed to issue implementing regulations within 180 days of the Act's effective date.
(c) 3 The RegisterRegisterThe term "Register" means the Register of Copyrights.Sec. 2(d)(5) shall establish and maintain a publicly available online database that contains each notice filed under subsection (a)(1).
Directs the Register of Copyrights to establish and maintain a publicly available online database containing all notices filed under the Act. This creates public transparency for copyrighted works used in training generative AI systems, but the obligation runs to the government, not to private parties.
(d)(1)–(6) ARTIFICIAL INTELLIGENCEArtificial IntelligenceThe term "Artificial Intelligence" means an automated system designed to perform a task typically associated with human intelligence or cognitive function.Sec. 2(d)(1).—The term "Artificial IntelligenceArtificial IntelligenceThe term "Artificial Intelligence" means an automated system designed to perform a task typically associated with human intelligence or cognitive function.Sec. 2(d)(1)" means an automated system designed to perform a task typically associated with human intelligence or cognitive function. (2) COPYRIGHTED WORKCopyrighted workThe term "copyrighted work" means a work protected in the United States under a law relating to copyrights.Sec. 2(d)(2).—The term "copyrighted workCopyrighted workThe term "copyrighted work" means a work protected in the United States under a law relating to copyrights.Sec. 2(d)(2)" means a work protected in the United States under a law relating to copyrights. (3) GENERATIVE AI MODELGenerative AI modelThe term "generative AI model" means a combination of computer code and numerical values designed to use Artificial Intelligence to generate outputs in the form of expressive material such as text, images, audio, or video.Sec. 2(d)(3).—The term "generative AI modelGenerative AI modelThe term "generative AI model" means a combination of computer code and numerical values designed to use Artificial Intelligence to generate outputs in the form of expressive material such as text, images, audio, or video.Sec. 2(d)(3)" means a combination of computer code and numerical values designed to use Artificial IntelligenceArtificial IntelligenceThe term "Artificial Intelligence" means an automated system designed to perform a task typically associated with human intelligence or cognitive function.Sec. 2(d)(1) to generate outputs in the form of expressive material such as text, images, audio, or video. (4) GENERATIVE AI SYSTEMGenerative AI systemThe term "generative AI system" means a software product or service that— (A) substantially incorporates one or more generative AI models; and (B) is designed for use by consumers.Sec. 2(d)(4).—The term "generative AI systemGenerative AI systemThe term "generative AI system" means a software product or service that— (A) substantially incorporates one or more generative AI models; and (B) is designed for use by consumers.Sec. 2(d)(4)" means a software product or service that— (A) substantially incorporates one or more generative AI modelsGenerative AI modelThe term "generative AI model" means a combination of computer code and numerical values designed to use Artificial Intelligence to generate outputs in the form of expressive material such as text, images, audio, or video.Sec. 2(d)(3); and (B) is designed for use by consumers. (5) REGISTERRegisterThe term "Register" means the Register of Copyrights.Sec. 2(d)(5).—The term "RegisterRegisterThe term "Register" means the Register of Copyrights.Sec. 2(d)(5)" means the Register of Copyrights. (6) TRAINING DATASETTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6).—The term "training datasetTraining datasetThe term "training dataset" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI model.Sec. 2(d)(6)" means a collection of individual units of material (including a combination of text, images, audio, or other categories of expressive material, as well as annotations describing the material) used to train a generative AI modelGenerative AI modelThe term "generative AI model" means a combination of computer code and numerical values designed to use Artificial Intelligence to generate outputs in the form of expressive material such as text, images, audio, or video.Sec. 2(d)(3).
Defines six key terms used throughout the bill: Artificial Intelligence, copyrighted work, generative AI model, generative AI system, Register, and training dataset. The definition of generative AI system is notable for its two-prong test — the system must both substantially incorporate one or more generative AI models and be designed for use by consumers, excluding enterprise-only or research-only systems.
(e) This Act shall take effect on the date that is 180 days after the date of the enactment of this Act.
The Act takes effect 180 days after enactment. This delayed effective date provides a window for the Register to issue implementing regulations and for covered persons to prepare initial notices for existing generative AI systems.