Key AI Legislation
Several legislative and regulatory measures have been proposed or enacted to address AI's impact on artists, writers, and creators. Below is a comprehensive breakdown of key developments:
U.S. Federal Initiatives
NO FAKES Act (2024)
Status: Proposed federal legislation
Provisions: Bans unauthorized AI-generated replicas of voices/likenesses in commercial, political, or deceptive contexts. Allows artists/estates to sue for damages (up to $5,000 per violation)
TRAIN Act
Status: Drafted (details pending)
Focus: Aims to regulate AI training data usage and protect creators from unauthorized exploitation of their works in AI models.
U.S. Copyright Office 2025 Report
Status: Policy guidance (non-legislative)
Key Rules:
AI-generated works without human input are uncopyrightable.
Copyright applies only to human-authored elements in hybrid works (e.g., edited AI outputs)1410.
Requires disclosure of AI use in registration applications.
​
U.S. Copyright Office 2025 Report
Status: Policy guidance (non-legislative)
Key Rules:
AI-generated works without human input are uncopyrightable.
Copyright applies only to human-authored elements in hybrid works (e.g., edited AI outputs)1410.
Requires disclosure of AI use in registration applications.
State-Level Laws
Tennessee ELVIS Act (2024)
Status: Enacted
Provisions:
Prohibits AI-generated voice/image replicas without consent.
Grants estates of deceased artists rights to sue.
California AB 2602 & AB 1836 (2024)
AB 2602: Bans unauthorized digital replicas in film/TV without explicit consent.
AB 1836: Extends posthumous likeness protections to estates.
Illinois AI Likeness Laws (Effective 2025)
Prohibits distributing AI-generated replicas of voices/images without permission.
Creates a private right of action for affected parties.
International Developments
UK’s AI Opportunities Action Plan (2025)
Focus: Balances AI innovation with copyright enforcement.
Key Proposals:
Transparency requirements for AI training data.
Exemptions for certain sectors in opt-out systems for copyrighted material.
EU AI Act (2023)
Status: Passed (implementation ongoing)
Relevance: Requires disclosure of AI-generated content and restricts deepfakes without consent (indirectly impacting creators).
Pending Proposals
Make it Fair Campaign (UK)
Demands stronger copyright enforcement, including tools to detect AI training data misuse and compensation mechanisms for artists.
Thomson Reuters v. ROSS Intelligence Precedent
Impact: Federal court ruled AI training on copyrighted materials without permission constitutes infringement, setting a judicial precedent for future cases.
Key Protections for Creators
NEXT:
AI-assisted works qualify only with substantial human creativity (e.g., editing, arranging, or transforming AI outputs)149.
Example: A painter modifying AI-generated sketches retains copyright over their modifications.
Laws in TN, CA, and IL protect against unauthorized AI replicas, with penalties up to $50,000 per violation.
Creators must rigorously document their input (prompts, edits) to prove human authorship in copyright claims.