Use-case guide

AI in Media & Entertainment

Generative AI hit media and entertainment first and hardest. The legal landscape consolidated rapidly: SAG-AFTRA + WGA 2023 strikes resulted in landmark AI clauses, the New York Times v. OpenAI litigation is testing training-data fair use, voice-cloning bans rolled out across states, and the EU AI Act layered Article 50 transparency on all generative content. The compliance pattern is now clear: consent + disclosure + provenance + payment. Every generative-AI feature in production now navigates this four-way intersection — and the platforms (Netflix, Disney, Spotify) have started requiring contractual flow-down of all four to vendor partners.

For: Studio executives, streaming platforms, music labels, generative-AI tool vendors for content, talent agencies, performer-rights counsel

What's at stake

Voice and likeness cloning is consent-gated by law in growing list of states

Tennessee ELVIS Act (2024), California AB 2602 + SB 1047 successor, NY S5959 — all establish that AI replicas of a person's voice or likeness require informed consent. Right of publicity is the underlying legal theory; statutory damages stack.

SAG-AFTRA and WGA contracts now have binding AI clauses

2023 SAG-AFTRA contract (and 2024 WGA MBA) require explicit performer consent + compensation for AI digital replicas, ban training-data use without consent, and require human credit. These contractual standards now propagate to non-union projects via industry custom.

EU AI Act Article 50 forces labelling of synthetic content

Generators must mark AI-generated audio, image, and video as machine-generated. Deployers in media must label deepfakes prominently. Article 4 also imposes AI-literacy duties on staff handling AI tools.

Training-data copyright litigation reshaping the field

NYT v. OpenAI, Getty v. Stability AI, Concord v. Anthropic — pending US/UK/EU cases will determine whether using copyrighted works to train generative models constitutes fair use / TDM exception. The downstream effect on commercial content production is enormous.

Regulations that apply

EU AI Act

Law
→

Article 50 labelling of AI-generated content. Article 52 deepfake disclosure. Article 53 GPAI provider duties incl. training-data transparency summaries.

Where in the text: Articles 50, 52, 53.

Tennessee ELVIS Act + state voice-cloning laws

Law
→

Right of publicity extended to voice + likeness. AI replicas require informed consent. Damages: actual + statutory + punitive.

Where in the text: Tenn. Code Ann. § 47-25-1101 et seq. (ELVIS Act).

California AB 1836 (Digital Replicas of Deceased Performers)

Law
→

Establishes that digital replicas of deceased performers require estate consent. Posthumous right of publicity extension.

Where in the text: Cal. Civ. Code § 3344.1.

SAG-AFTRA + WGA AI clauses (contractual)

Law
→

Industry-de-facto standard. Performer-consent for digital replicas; bargained compensation; restrictions on training-data use without consent.

Where in the text: SAG-AFTRA 2023 TV/Theatrical Contract Art. 64; WGA MBA 2023 Art. 72.

Do

  • ✓Get layered consent for any voice/likeness use: the original performance, the specific synthetic use, the duration, and the territory. Lifetime/all-purpose consent is unenforceable in most jurisdictions.
  • ✓Label AI-generated content visibly in the playback surface AND via C2PA/SynthID metadata. The EU AI Act + state deepfake laws + platform contracts all point at the same thing.
  • ✓Document training-data provenance for any model used in commercial productions. The 'fair use' question is contingent on facts you need to be able to produce.
  • ✓Build an AI-use disclosure in script credits, music credits, and metadata. Industry standards (SAG-AFTRA, IFPI, RIAA) are converging on visible attribution.
  • ✓Maintain a consent-revocation pathway. Several states require it; talent contracts increasingly demand it.

Don't

  • ✗Don't use scraped voice or video of public figures to train commercial generators without licensing — even where the input data is 'publicly available'.
  • ✗Don't pass off AI-generated performances as human-performed in EU markets. Article 50 makes that an enforcement-grade omission.
  • ✗Don't 'digitally double' a performer for use beyond the contracted scope. SAG-AFTRA grievance + state right-of-publicity damages stack here.
  • ✗Don't deploy a deepfake of an existing celebrity in advertising without that person's consent — multiple US states have made this criminal.
  • ✗Don't omit a synthetic-content disclosure on platforms that increasingly require it (TikTok, YouTube, Instagram all rolled out 2024 policies).

Also worth knowing

For music platforms: PROs (BMI, ASCAP, GEMA) are developing AI-generated-music registration policies; submission rules are tightening. For animation studios: the line between 'AI-assisted production' (workflow tools) and 'AI-generated content' (the work itself) drives different SAG-AFTRA + WGA exposure. For video game studios: generative AI in NPC dialogue interacts with both Article 50 disclosure and SAG-AFTRA interactive contracts.

Want a tailored answer?

The wizard takes your jurisdiction, AI use case, and data types and gives you the top-3 regulations to focus on — in 60 seconds.

Start the wizard →

Educational guide. Not legal advice. For specific compliance decisions, consult qualified counsel in the relevant jurisdiction.

Note: this guide was drafted with AI assistance — Anthropic Claude.