AI in Entertainment Production: Legal Issues for Creators

Artificial intelligence has moved from an emerging technology to a daily production tool. Film editors use AI for rotoscoping and color grading. Music producers generate beats, stems, and vocal processing with AI-powered software. Content creators use generative AI for scripts, thumbnails, social media copy, and visual effects. Post-production studios integrate AI into noise reduction, upscaling, audio cleanup, and automated editing workflows. The creative applications are expanding rapidly, but the legal framework governing AI-generated content is still catching up.

For entertainment professionals in New York, understanding the legal issues around AI in production is not optional. Copyright ownership, contract provisions, union and guild obligations, licensing, and liability all intersect when AI tools are part of the creative process. This guide covers what you need to know to use AI tools in your productions while protecting your rights, your clients' rights, and your business.

Copyright and AI-Generated Content

The Human Authorship Requirement

U.S. copyright law requires human authorship. The Supreme Court declined to review the D.C. Circuit's ruling in Thaler v. Perlmutter in March 2026, confirming that works created solely by AI systems cannot be copyrighted. The Copyright Office has consistently held that AI-generated material, without meaningful human creative input, is not eligible for copyright registration. This means that if you use an AI tool to generate an image, a piece of music, or a video clip, and the AI made the creative decisions about the expressive elements, the output is not protectable by copyright. Anyone can use it, copy it, or build on it without your permission.

AI-Assisted vs AI-Generated

The critical distinction is between AI-generated content (where the AI determines the expressive elements) and AI-assisted content (where a human exercises meaningful creative control over the output). AI-assisted works can be copyrighted, but only the human-authored elements receive protection. The Copyright Office evaluates the degree of human creative control on a case-by-case basis. Factors that strengthen your copyright claim include selecting and arranging AI outputs in a creative manner, substantially editing, modifying, or transforming the AI output, combining AI-generated elements with human-created elements, exercising iterative creative direction over the AI tool, and making subjective creative decisions about the final product. The more creative control you exercise, the stronger your copyright claim. Simply entering a text prompt and accepting the output, without further creative modification, is unlikely to satisfy the human authorship requirement.

Copyright Registration for AI-Assisted Works

When applying to register a copyright for a work that includes AI-generated material, you must disclose the AI-generated components and identify the human-authored elements. The Copyright Office will evaluate whether the human contribution is sufficient to support registration. Failure to disclose AI involvement can result in cancellation of the registration. If you use AI tools in your creative workflow, document your process: save prompts, drafts, intermediate versions, and notes about the creative decisions you made. This documentation supports your copyright claim and demonstrates the human authorship required for protection. For more on copyright registration, see our copyright registration guide.

AI in Pre-Production

AI tools are increasingly used in pre-production for scriptwriting assistance (generating dialogue, scene descriptions, or story outlines), storyboarding (generating visual representations of scenes from text descriptions), casting research (using AI to analyze performance styles and audience demographics), location scouting (AI-powered image analysis and matching), and budgeting and scheduling (AI-driven production planning tools). When AI is used in pre-production, the key legal question is who owns the output. If the AI-generated material is incorporated into the final production, the copyright implications described above apply. Scripts, storyboards, and other pre-production materials generated primarily by AI may not be copyrightable, which means they cannot be protected from unauthorized use. If an AI-generated script outline is not copyrightable, another production company could use a similar outline without infringing your rights.

AI in Production and Post-Production

In production and post-production, AI tools are used for visual effects (AI-generated backgrounds, characters, and environments), de-aging and digital doubles (creating or modifying performer appearances), audio processing (noise reduction, dialogue enhancement, automated mixing), color grading (AI-assisted color matching and correction), editing (automated assembly cuts, scene detection, content-aware editing), music composition (AI-generated scores, sound design, and audio beds), and subtitling and translation (automated captioning and language translation).

Each of these applications raises specific legal questions. AI-generated visual effects that create recognizable likenesses of real people implicate the right of publicity. AI tools that process or modify existing copyrighted content (such as upscaling a copyrighted film clip or using AI to remix a copyrighted song) may constitute derivative works requiring a license from the original rights holder. AI-generated music used in a production may not be copyrightable, creating gaps in your ability to control how the final product is used downstream.

Performer Rights and Digital Replicas

The use of AI to create digital replicas of performers, whether through deepfakes, de-aging technology, or AI-generated voice cloning, raises significant legal issues under both contract law and the right of publicity. New York's right of publicity law (Civil Rights Law Sections 50 and 51) protects individuals from the unauthorized commercial use of their name, portrait, picture, or voice. Using AI to create a digital replica of a performer without their informed consent violates these protections, regardless of how realistic the replica is.

The SAG-AFTRA contract negotiated in 2023 established groundbreaking protections for performers in the context of AI. Key provisions include a requirement for clear and conspicuous informed consent before creating a digital replica of a performer, fair compensation for the use of a performer's digital replica, the right to approve or reject specific uses of the digital replica, restrictions on using AI to replace or reduce performer work without negotiation, and protections for background performers whose likenesses are captured and could be replicated by AI. Producers working with union talent must comply with these provisions. Non-union productions should adopt similar protections as a matter of best practice and legal risk management, as state right of publicity laws apply regardless of union status.

Contract Provisions for AI Use

Every production contract should now address AI use explicitly. Whether you are drafting a client agreement, a talent contract, a crew deal memo, or a vendor agreement, the following provisions should be considered: a definition of AI tools (specifying which tools are permitted and which are prohibited), ownership of AI-assisted outputs (who owns the elements generated with AI assistance), disclosure obligations (requiring parties to disclose when AI tools are used in the creation of deliverables), representations of originality (confirming that the deliverables do not infringe third-party rights, including rights that may be implicated by AI training data), indemnification (each party's obligation to defend and compensate the other for IP claims arising from AI use), credit and attribution (how AI use affects screen credits, writing credits, and other professional attributions), and compliance with union and guild requirements (confirming that AI use complies with applicable collective bargaining agreements).

For production companies that provide services to clients, the production services agreement should specify whether the client has approved the use of AI tools and what representations the production company makes about the originality and copyrightability of the deliverables. A client who pays for a commercial video may not realize that AI-generated elements of the video are not copyrightable and cannot be protected from unauthorized use. Failing to disclose this creates liability risk for the production company. For more on production contracts, see our production agency legal issues guide.

AI and Music Production

AI music tools (such as Suno, Udio, AIVA, and Amper) can generate complete compositions, including melody, harmony, arrangement, and even vocals. For music producers and composers, the legal issues include copyright ownership of AI-generated compositions (which may not be protectable if the AI made the creative decisions), sync licensing complications (licensees may require representations that the music is copyrightable, which AI-generated music may not satisfy), royalty collection (performing rights organizations may not accept AI-generated works for registration if they lack human authorship), and training data liability (if the AI tool was trained on copyrighted music without authorization, the output could be subject to infringement claims by the original rights holders).

Music producers who use AI tools should treat the AI output as raw material and apply substantial human creative input (arrangement, editing, mixing, instrumentation choices, lyric writing) to strengthen the copyright claim on the final work. Document every creative decision you make beyond the initial AI generation. For more on music contracts and rights, see our music contracts guide and our music rights guide.

Training Data and Infringement Risk

A separate but related legal issue is whether AI companies can use copyrighted works to train their models without permission. Multiple major lawsuits are pending, and federal courts have reached different conclusions. In one notable case, a court found that training an AI legal research tool on copyrighted headnotes was not fair use. In another, training on lawfully acquired books was found to be fair use, though using pirated copies was not. The legal landscape is evolving, and the outcome of these cases will affect the liability exposure of both AI tool providers and the creators who use their outputs.

For entertainment professionals, the practical implication is that if your AI-generated output is substantially similar to a specific copyrighted work, you could face an infringement claim regardless of your intent or the AI tool's role in creating the output. Using reputable AI tools with transparent, licensed training data reduces this risk but does not eliminate it. When possible, verify that the AI tool provider has appropriate licenses or fair use defenses for its training data, and maintain documentation of your creative process to demonstrate the human authorship of your final work.

Protecting Your Work Product: Data Privacy and AI Training

One of the most pressing concerns for production companies and creators using AI tools is what happens to the content they upload. When you feed raw footage, audio files, scripts, character designs, or other proprietary material into an AI platform for processing, editing, or enhancement, the critical question is whether that content becomes part of the platform's training data, accessible to other users or the platform itself for future model improvement.

How AI Models Learn from User Inputs

Most AI tools operate on large language models (LLMs) or generative models that were trained on massive datasets of existing content. But many platforms also use customer inputs to improve their models on an ongoing basis. When you upload a video clip to an AI editing tool, submit a script to an AI writing assistant, or process audio through an AI mixing platform, the platform may retain your content, analyze it, and incorporate patterns from it into future model updates. Once your creative material enters the training pipeline, it can influence outputs that the platform generates for other users. Your unreleased film footage, your proprietary sound design, or your unpublished script concepts could, in theory, inform content generated for a competitor using the same platform.

Reading the Terms of Service

Every AI platform's terms of service contain provisions governing how your uploaded content is used. The key provisions to review include the data retention policy (how long the platform retains your uploaded content after processing), the training data clause (whether the platform uses your inputs to train or improve its models), the license grant (what rights you are granting the platform when you upload content, and whether that license is broad enough to include training use), the opt-out mechanism (whether you can opt out of having your content used for training, and whether the opt-out is available by default or requires affirmative action), and the data deletion policy (whether you can request deletion of your uploaded content and whether deletion removes it from future training datasets). Some platforms offer enterprise or professional tiers that contractually exclude customer inputs from training data. Others include training rights in their standard terms and require users to opt out. A few platforms use all uploaded content for training with no opt-out option. Before uploading any proprietary content to an AI platform, review the terms of service with your attorney and confirm in writing how your content will be handled.

Confidentiality and NDA Conflicts

Production companies routinely operate under non-disclosure agreements with clients, studios, distributors, and talent. If you upload content that is subject to an NDA into an AI platform that uses customer inputs for training, you may be violating your confidentiality obligations. The NDA likely defines confidential information broadly enough to cover production materials, and uploading those materials to a third-party platform that retains and reuses them is a disclosure that the NDA prohibits. Before using any AI tool on a client project, verify that your use is consistent with the confidentiality provisions in your client agreement. If the AI platform's terms do not guarantee that your content will not be used for training or shared with third parties, using the platform on confidential client materials creates both a breach of contract risk and a reputational risk that could cost you the client relationship.

Enterprise Agreements and Data Isolation

For production companies that rely heavily on AI tools, negotiating an enterprise agreement with the AI provider is the safest approach. Enterprise agreements can include contractual guarantees that customer inputs will not be used for model training, data isolation provisions ensuring your content is processed in a separate environment, data deletion requirements specifying how and when your content is purged from the platform's systems, audit rights allowing you to verify compliance with the data handling provisions, indemnification from the AI provider for breaches of the data handling obligations, and compliance with applicable data protection regulations (including the EU's GDPR if you work with European clients or talent). The cost of an enterprise agreement is justified when the alternative is exposing proprietary client content to uncontrolled use by a third-party AI platform.

On-Premise and Local AI Solutions

An alternative to cloud-based AI tools is running AI models locally, on your own hardware or within your own network infrastructure. Local deployment eliminates the data privacy risk entirely because your content never leaves your systems. Several AI tools used in production and post-production, including certain video editing, color grading, audio processing, and visual effects applications, offer local installation options that do not require uploading content to external servers. Local deployment requires more computing resources and technical expertise, but for productions handling highly sensitive content (unreleased films, confidential commercial campaigns, content involving minors, or content subject to strict NDAs), the privacy benefits outweigh the cost and complexity. Your IT team and your attorney should evaluate the trade-offs between cloud and local AI deployment for each project based on the sensitivity of the content and the contractual obligations governing it.

Insurance Considerations

Standard production insurance policies, including errors and omissions (E&O) coverage, may not cover claims arising from AI-generated content. Insurers are still developing policies that address AI-specific risks, including copyright infringement claims based on AI training data, right of publicity claims arising from AI-generated digital replicas, and professional liability claims based on the use of AI tools that produce infringing or defamatory output. Before relying on AI tools in a production, review your E&O policy with your broker to determine whether AI-related claims are covered, excluded, or subject to specific conditions. If coverage is unclear, request a policy endorsement that explicitly addresses AI use in your productions. For more on production insurance, see our production agency legal guide.

Best Practices for Using AI in Production

To minimize legal risk while leveraging AI tools in your creative workflow, adopt these practices: document your creative process thoroughly (save prompts, intermediate versions, editing decisions, and notes about creative choices), exercise meaningful creative control over AI outputs (do not use raw AI output as a finished product), disclose AI use in copyright applications, contracts, and to clients, review and comply with applicable union and guild provisions, use AI tools from providers with transparent and defensible training data practices, obtain informed consent before creating digital replicas of performers, update your production contracts to address AI use explicitly, review your insurance coverage for AI-related claims, and consult your entertainment attorney when using AI in ways that affect copyright ownership, performer rights, or client deliverables. For more on entertainment law, visit our entertainment law practice page.

Frequently Asked Questions