Synthetic Performers & Intellectual Property Rights: The Battle for Likeness, Residuals, and the Future of Entertainment Labor

Executive Summary

The September 2025 debut of Tilly Norwood—a fully synthetic actress created by the studio Xicoia—triggered what appears to be a decisive victory for entertainment labor unions: the synthetic performer was immediately rejected by major talent agencies following public outcry from stars, director opposition, and SAG-AFTRA’s unequivocal legal condemnation. Yet in the same month, Xania Monet, an AI-generated musician created via the Suno platform, signed a $3 million record deal with Hallwood Media and charted on Billboard.

This divergence reveals a fundamental asymmetry in the entertainment industry’s capacity to resist synthetic performer adoption. The institutional strength of SAG-AFTRA—bolstered by contractual protections negotiated in the 2023 strike and codified in 2025 commercials, video game, and interactive media agreements—erected a firewall against film industry synthetic performers. Simultaneously, weaker union structures in music, combined with streaming services’ pre-existing elimination of residual payments, created a permissive regulatory environment for Xania’s success.

The broader threat, however, is not a sudden wholesale replacement of human actors by synthetics, but rather a gradual bifurcation of the entertainment labor market into a “premium human tier” (established stars commanding scarcity premiums) and a “synthetic utility tier” (background actors, voice actors, commercial presenters displaced into obsolescence). This restructuring will disproportionately harm entry-level and middle-tier performers while concentrating income among the few, fundamentally disrupting the financial viability of the actor profession and decimating the residual income model that has sustained creative labor for decades.

---

1. The Tilly Norwood Incident: Why the Firewall Held

The Pitch & Immediate Backlash

On September 27, 2025, at the Zurich Film Festival Summit, Xicoia founder Eline Van der Velden announced the launch of Tilly Norwood, a fully synthetic actress—not a digital double of a specific human, but a composite character generated entirely by AI, trained on performances of “countless professional performers” without permission or compensation.​

Van der Velden positioned Tilly as a breakthrough: “We believe the next generation of cultural icons will be synthetic: stars who never tire, never age, and can interact with their fans.” She announced that Tilly was in negotiations with major talent agencies, and Xicoia planned to develop a universe of “over 40 digital personalities.”​

The industry response was swift and uniformly hostile:

• Emily Blunt told Variety: “Does it disappoint me? That’s terrifying. Good Lord, we’re screwed. That is really, really scary. Come on, agencies, don’t do that. Please stop taking away our human connection.”​
• Whoopi Goldberg on The View: “The problem with this, in my humble opinion, is that you are suddenly up against something that’s been generated with 5,000 other actors. That’s a lot of competition.”​
• Abigail Breslin posted: “I beg every actor I know to [please] boycott this. If your agency is trying to sign an AI actor, you should leave them. It’s such a fuck you to the entire craft.”​

SAG-AFTRA’s Legal Response

SAG-AFTRA issued a statement that proved legally dispositive. The union declared:​

“Tilly Norwood is not an actor; it’s a character generated by a computer program that was trained on the work of countless professional performers—without permission or compensation. It has no life experience to draw from, no emotion… Using stolen performances to put actors out of work jeopardizes performer livelihoods and devalues human artistry. Signatory producers should be aware that they may not use synthetic performers without complying with our contractual obligations, which require notice and bargaining whenever a synthetic performer is going to be used.”

This statement did two things: (1) Reframed Tilly from a technical achievement to a copyright infringement—the AI was “trained on stolen work”; and (2) asserted that use of any synthetic performer on union productions requires negotiation and compensation, effectively making Tilly legally unusable in 90% of theatrical/TV film production (where SAG-AFTRA signatory rates apply).​

Talent Agency Retreat

Following the public backlash, major talent agencies—initially rumored to be negotiating with Xicoia—publicly disassociated. Gersh and other large agencies feared a revolt from their A-list rosters if they signed Tilly. Van der Velden’s own defensive response—positioning AI as “a new tool, a new paintbrush”—fell flat; the Instagram post announcing Tilly received thousands of negative comments before the account disabled further interaction.​

By early October 2025, Tilly Norwood had effectively been blacklisted from the entertainment industry. No major agencies signed her. No signatory producers announced projects using her likeness. The synthetic actress became a case study in the power of unionization and contractual enforcement.

---

2. The Xania Monet Paradox: Why Synthetics Succeeded in Music

While Tilly Norwood was being rejected in film, a parallel story unfolded in music with radically different results.

The Artist & Record Deal

Xania Monet emerged in September 2025 as an AI-generated R&B/Gospel artist created through a collaboration between Telisha “Nikki” Jones (Mississippi-based poet and designer) and Suno (an AI music generation platform). Jones retained creative control—writing lyrics and directing Monet’s artistic direction—while Suno generated the vocals, production, and visual persona.​

The results were commercially impressive:

• Debut single “How Was I Supposed to Know” achieved #1 on the R&B Digital Song Sales chart​
• “Let Go, Let God” charted #21 on Billboard Hot Gospel Songs; #25 on Emerging Artists​
• Nearly 10 million streams, 465,000 monthly Spotify listeners​
• $50,000+ in revenue from a 5-song catalogue within months​

In September 2025, Hallwood Media (with executive producer Neil Jacobson, formerly of Interscope) signed Xania Monet to a $3 million record deal after a competitive bidding war.​

Why Music Accepted What Film Rejected

The divergence between Tilly’s rejection and Xania’s success stems from four structural differences:

1. Weaker Union Enforcement in Music

• Film: SAG-AFTRA (160,000+ members) negotiates all theatrical/TV agreements; signatory producers bound by strict contract terms
• Music: American Federation of Musicians (AFM) and SAG-AFTRA’s music/recording divisions have less leverage over streaming platforms and independent labels
• Streaming impact: Music streaming platforms (Spotify, Apple Music) treat AI-generated tracks identically to human-created music; no contractual requirement to use human performers​

2. Pre-Existing Elimination of Residuals in Streaming

• Film: Residuals remain a core compensation mechanism (though declining); union fighting to preserve them
• Music: Streaming services already decimated traditional music industry residuals in the 2010s; flat fees or per-stream micropayments became standard
• Result: Xania’s synthetic nature added less incremental harm to an already-fractured income model​

3. Clearer Ownership Model

• Tilly Norwood: Created by Xicoia (ownership unclear); trained on “countless professional performers” without compensation (copyright infringement red flag)
• Xania Monet: Telisha Jones owns creative/lyrical rights; Suno provides platform/voice generation; revenue split negotiated upfront (clearer IP chain of title)​

4. Brand Risk Differential

• Film/Major Studios: Reputational risk of public backlash; A-list talent threatening departures if studios use synthetics
• Music/Independent Labels: Neil Jacobson (Hallwood Media) positioned AI as “the future of our medium”; less entrenched talent roster to placate; precedent-setting seen as competitive advantage​

Industry Division Over Xania’s Success

Unlike the near-universal condemnation of Tilly, Xania’s deal produced mixed reactions:

Support:

• Timbaland: Praised Xania’s music; advises Suno directly; launched his own AI music label, Stage Zero, signing virtual artist TaTa​
• Neil Jacobson: “AI is the future of our medium”—signaling early mover advantage​

Opposition:

• Grammy artists Kehlani & SZA: Criticized AI music as devaluing human artistry; raised concerns about unauthorized training data (Suno’s voice model trained on unknown corpus)​
• Recording Industry Association of America (RIAA): Filed lawsuits against Suno/Udio alleging copyright infringement for training AI on copyrighted music without permission​

The unresolved question: If Xania’s voice was trained on samples of real singers (without compensation), does she constitute a synthetic performer whose creation violated Right of Publicity rights?

---

3. The Legal Firewall: California Laws & Federal Pending Legislation

California AB 2602: Contract Protection Against Unfair Terms

Effective January 1, 2025, California Assembly Bill 2602 renders unenforceable any contract provision that:

1. Permits digital replica use to substitute for in-person work
2. Fails to describe intended uses with “reasonably specific” detail
3. Individual not represented by legal counsel or applicable labor union​

Practical impact: Studios cannot use vague language like “all media known or hereafter devised” to claim blanket rights to create digital replicas. A performer must either:

• Negotiate specific, written agreement listing exact uses (e.g., “commercials for automotive products only”)
• Be represented by legal counsel or union during negotiation
• Refuse to sign without specificity​

Critical ambiguity: What constitutes “reasonably specific”? A court could interpret “audio-visual works” as sufficiently specific (allowing broad use) or insufficiently specific (requiring detailed delineation of each medium/product). This threshold remains undefined, creating litigation risk.​

Gap: Training data use not addressed. AB 2602 prohibits unfair contracts, but does not prohibit the training of AI systems on performer footage without consent—only the commercial use of replicas created thereafter.​

California AB 1836: Posthumous Rights of Publicity

Effective January 1, 2025, AB 1836 prohibits creating or distributing digital replicas of deceased personalities’ voices or likenesses without prior consent from their estates.​

Example blocked: Creating a digital replica of Michael Jackson to perform a new song or endorsement commercial would require explicit consent from the Jackson estate—not just general publicity rights.​

Penalties: Greater of $10,000 or actual damages​

Exceptions: Fleeting uses, documentaries, news, criticism, satire, parody allowed (First Amendment carve-out)​

Federal NO FAKES Act (Pending, January 2026)

The “Nurture Originals, Foster Art, and Keep Entertainment Safe Act of 2025” (H.R. 2794 / S.1367) was reintroduced April 9, 2025, and remains in House/Senate Judiciary committees as of January 2026.​

Status: Bipartisan support (Rep. Salazar [R-FL-27], Sen. Coons [D-DE]); revised version (now 39 pages) incorporates feedback from Google, OpenAI, record labels, SAG-AFTRA​

Key Provisions:

A. Digital Replication Right (Core Innovation)

• Creates federal property right in “voice or visual likeness in digital replicas”
• During lifetime: Non-assignable, licensable (exclusive/non-exclusive)
• License duration: Maximum 10 years while person living
• For minors: Maximum 5 years or until age 18 (whichever earlier); requires court approval
• License requirement: Written agreement with “reasonably specific description” of uses​

B. Post-Mortem Rights (Controversial)

• Transfer: Transferable to executors/heirs for 10 years
• Renewal: Renewable for 5-year periods if “active and authorized public use” demonstrated (2-year window required)
• Maximum duration: 70 years from death
• Registration: Copyright Office maintains directory of post-mortem rights; renewal filings required​

Criticism: Creates financial incentive to commercialize the dead; unfairly favors commercialized estates (70 years) over non-commercialized estates (10 years); burdens Copyright Office with administrative registration​

C. Liability Structure (DMCA-like Safe Harbor)

• Unauthorized distribution: $5,000-$750,000 per work depending on violator type
• Tools designed for unauthorized replication: $5,000-$750,000 per product/service
• Online services: Reduced liability ($25,000) if they register designated agent, implement notice-and-takedown, use digital fingerprinting to prevent re-uploads​
• No monitoring duty: Services not required to proactively police content; only act upon notice​

D. Enforcement Mechanism

• Right holder, eligible plaintiffs, or sound recording artists with exclusive contracts can sue
• 3-year limitation period from discovery of violation
• Subpoena power to compel online services to disclose alleged infringers’ identity​
• Disclaimer not a defense: Company cannot claim “unauthorized” deepfake to escape liability​

Synthetic Performer Timeline: Regulatory vs. Industry Adoption (2024-2026) 

Critical Gaps & Ongoing Debate

1. “Reasonably Specific” Ambiguity Persists
Both AB 2602 (California state law) and NO FAKES Act (federal) require “reasonably specific” description of digital replica uses. Neither defines this threshold, creating litigation risk.​

Professor Pam Samuelson argues the standard could be dangerously vague: Is “audio-visual works” reasonably specific? “Promotions”? “All platforms”? Without greater clarity, performers could sign contracts permitting pornographic, defamatory, or otherwise abusive digital replicas without explicit awareness.​

2. Post-Expiration Reuse Loophole
NO FAKES Act allows licensing of digital replicas for 10-year terms, but permits continued use of replicas created during that period after expiration. This undermines the durational limit—a performer’s replica created in year 10 could be used indefinitely after year 10 expires.​

3. Training Data Liability Unresolved

• California AB 2602: Doesn’t address whether training AI on performer footage without consent is legal
• Federal NO FAKES Act: Doesn’t retroactively criminalize or penalize AI training on copyrighted performances
• RIAA lawsuits: Challenge whether Suno/Udio violated copyright by training on copyrighted music, but outcome uncertain​
• Question: If Xania Monet’s voice was trained on thousands of singers’ recordings, do those singers have recourse?

4. Preemption Uncertainties (Federal Act)
NO FAKES Act includes preemption of state laws BUT excludes:

• State laws “existing as of January 2, 2025 regarding a digital replica” (California laws grandfathered)
• State laws on sexually explicit digital replicas
• State laws on election-related deepfakes

Ambiguity: Does “regarding a digital replica” preempt broader state right of publicity laws that protect digital replicas but weren’t specifically drafted with AI in mind? This creates a “patchwork of uncertainty” that may invite regulatory arbitrage—producers routing synthetic talent through states with weaker protections.​

5. Minors & Representative Authorization Risk
NO FAKES Act allows “authorized representatives” (parents, agents, managers) to license a minor’s digital replica without direct minor consent. Post-age-18, minors are bound to contracts their representatives signed. This creates risk of coercive situations where young athletes/models are locked into broad digital replica licenses signed years earlier.​

---

4. Union Protections: The Contractual Arsenal (2023-2025)

SAG-AFTRA 2023 Theatrical/TV Agreement

Following the 2023 strike, SAG-AFTRA negotiated the first binding digital replica protections:

• Digital Replica Use: Requires informed consent from original performer
• Separate Compensation: Producer must pay performer for digital replica use (separate from initial performance fee)
• Digital Alterations: Specific procedures required when producers alter previously recorded performer footage​
• Replica Studios Agreement: Two-tier structure for voice replicas: “Development” (internal creation) and “Licensing/External Use” (commercial deployment)​
• Synthetic Performers Definition: Characters not recognizable as specific individuals don’t require individual consent​

Impact on Tilly Norwood: Under this agreement, any SAG-AFTRA signatory producer using Tilly must:

1. Notice the union (required bargaining)
2. Negotiate compensation (residual equivalent or buyout)
3. Comply with residual payment obligations if Tilly is used in place of a human actor

This contractual obligation made Tilly legally unusable on union productions, effectively blacklisting her from mainstream theatrical/TV film.

SAG-AFTRA 2025 Commercials Contract (Ratified May 2025)

The most recent major agreement (effective 2025-2028) established detailed AI compensation:

• Digital Replica Session Fee: 1.5x SAG scale per commercial where replica used​
• Use & Holding Fees: Calculated as if performer worked live (on-air fees, holding fees, etc.)​
• **Synthetic Performer (Mixed): **Session payment + applicable use fees at 1.5x scale​
• Synthetic Performer (Exclusive): Per-hour minimum TBD (only when no humans used)​
• Payment timing: Within 30 days of airdate​
• Wage increases: 5% (year 1), 4% (year 2), 3% (year 3)​
• Pension & Health contribution: 23.5%​

Economic impact: 2025 deal totals $218.4 million in new performer earnings and benefits over three years—nearly double the $120 million gains from the 2022 agreement.​

The 1.5x scale premium for digital replicas was designed to make synthetic casting economically unattractive (why use a synthetic if you have to pay 1.5x anyway?), effectively pricing synthetic performers out of union productions.​

SAG-AFTRA 2025 Video Game Agreement (June 2025, 95.04% Approval)

Notably, the video game industry agreement went further than film/TV protections:

• Digital replica rates: Comparable to direct performer work (not discounted)​
• Consent & disclosure requirements: Mandatory; cannot create replica without informed consent​
• Strike protection: Performers can suspend consent for replica use during strikes​
• Intent: Prevent studios from circumventing union by creating replicas from non-covered work​

Key phrase: “Performers can suspend consent during strikes”—ensuring synthetic performers cannot be weaponized to break labor actions.

UK Equity (December 2025): Industrial Action Vote

Britain’s equity union took the most aggressive stance yet. On December 17, 2025, Equity balloted 7,000+ performers (the largest such ballot in union history) on industrial action related to AI protections.​

• Result: 90%+ voted to refuse digital scanning on set​
• Target: Negotiations with Pact (Producers Alliance for Cinema & Television)​
• Unresolved issue: AI training data use (recorded performances, digital scans) remains disputed​
• Threat: Statutory ballot for strike if negotiations don’t produce “future-facing protections that extend beyond established safeguards” negotiated elsewhere​

Significance: This is not about synthetic performers yet, but about the foundational permission to scan performers’ likenesses for AI training. Equity is effectively trying to prevent studios from capturing performer biometrics without explicit compensation.​

---

5. The Residual Erosion: From Safety Net to Anachronism

The deepest threat posed by synthetic performers is not job displacement per se, but the destruction of the residual income model that has sustained creative labor for decades.

The Traditional Residual Economy

Historically, residuals have functioned as a safety net:

• Source: Each time a film/TV show is re-aired, rerun, or resold (DVD, syndication, streaming licensing)
• Frequency: Actors earn additional compensation for each re-use
• Duration: Decades of long-tail income from successful works
• Pool structure: Residual fees feed union pension & health funds, providing healthcare and retirement for the entire membership​

Example: An actor earning $5,000 for a theatrical role might receive:

• Initial fee: $5,000
• Residuals (theatrical re-release): $1,000
• Residuals (TV syndication, 10 runs): $2,000
• Residuals (DVD/Blu-ray sales): $500
• Total over 20 years: $8,500 (70% additional income)

Streaming’s Elimination of Residuals

Streaming platforms fundamentally broke this model beginning in the 2010s:

• No reruns: Streaming is on-demand; shows don’t “run” at scheduled times
• Flat-fee model: Netflix, Amazon, Apple TV, Disney+ pay upfront lump sums; no per-view metrics
• Revenue opacity: Platforms don’t publish viewing metrics, making residual calculation impossible
• Labor devaluation: SAG-AFTRA unable to negotiate residuals in 2017 Netflix deal; actors shifted to flat-fee dependency​

Result: Median actor residual income declined 68% between 2010-2023 as streaming dominance increased.​

Synthetic Performers Accelerate the Death Spiral

The introduction of synthetic performers threatens the remaining residual pool in two ways:

1. Direct Displacement Erodes Union Pension Contributions

• SAG-AFTRA Pension & Health Fund receives contributions from every union member’s paycheck
• If 10% of background/utility roles replaced by synthetics, union contributions drop accordingly
• Remaining members pay higher percentages to fund fixed healthcare/retirement obligations
• Cascading effect: Reduced income → fewer members → higher per-capita fund burden → pension insolvency risk​

2. Coercive Bargaining Shifts Power to Producers
The existence of synthetic alternatives creates a BATNA (Best Alternative to Negotiated Agreement) for producers:

• Actor demands: “I want residuals for streaming re-use”
• Producer response: “If you don’t accept buyout, we use a synthetic”
• Result: Buyout prevalence increases​

Data: UK AV performer survey found 96% of performers perceive buyout contracts increasing, suggesting synthetic threat is already triggering coercive buyout pressure.​

---

6. Performer Vulnerability Matrix: Asymmetric Displacement Risk

Performer Job Security Impact: Synthetic Performer Displacement Risk by Category (2026-2028) 

The threat of synthetic performer adoption is not uniform; it follows a clear pattern of labor precarity:

High-Risk Categories (30-50% displacement risk by 2028):

• Background actors: No dialogue, easily replaceable, no union minimum negotiating power
• Voice actors: Commercial/dubbing market; ElevenLabs/similar already replace 30%+ of voice-over work
• Commercial presenters: Brands prefer synthetics (no scandal risk, full creative control, lower cost)
• Stunt performers: High risk for repetitive stunts (falls, crashes); hybrid model emerging
• Entry-level roles: Path-to-stardom shortened; fewer “break” roles for unknowns

Medium-Risk (10-25% displacement risk):

• TV actors in scripted episodic: Protected by union, but budget pressures may force hybrid models (synthetic secondary characters)

Low-Risk (5% displacement):

• Established film stars: Scarcity premium emerging; “real” actors positioned as luxury brand
• A-list talent: Union protection + star power + talent shortage ensures continued demand

The Bifurcated Labor Market

By 2028, the entertainment industry is likely to stratify into:

Premium Human Tier:

• Established actors commanding 20-50% fee premiums for “guaranteed human performance”
• Digital double licensing as passive income stream (strict approval rights)
• Estimated 5,000-10,000 actors globally; ~2% of profession

Synthetic Utility Tier:

• Background actors, voice actors, commercial presenters largely displaced
• Remaining roles filled by synthetics (no residuals, health, pension)
• Estimated 50,000-100,000 performers displaced; ~15-25% of US acting workforce

Precarious Middle Tier:

• Mid-tier TV/film actors in hybrid productions (some scenes human, some synthetic)
• Buyout pressure increases; residual income further erodes
• Estimated 200,000-300,000 performers; vulnerable to downward mobility

---

7. Why Synthetic Performers Failed in Film But Succeeded in Music

Tilly Norwood vs. Xania Monet: Why Synthetic Film Actress Failed, Synthetic Musician Succeeded 

The divergence between Tilly Norwood’s rejection and Xania Monet’s $3 million deal illuminates the structural conditions that enable synthetic performer adoption:

Union Strength

• Film: SAG-AFTRA (160,000+ members) negotiates binding contracts; signatory producers cannot use synthetics without compensation
• Music: AFM/Recording labels less coordinated; independent labels not bound by union terms
• Winner: Music (less enforcement, more freedom)

Residual Economics

• Film: Residuals still matter (declining but present); synthetic use threatens union pension fund
• Music: Residuals already eliminated; synthetic synthetic performer adds minimal incremental harm
• Winner: Music (existing residual loss makes synthetics less objectionable)

Public Backlash

• Film: A-list stars (Blunt, Goldberg, Breslin) publicly opposed; agency threat to cancel rosters
• Music: Mixed reaction; Timbaland endorsement; no unified industry opposition
• Winner: Music (dispersed opposition)

Training Data Liability

• Tilly: Created from “stolen work” (copyright infringement red flag)
• Xania: Clearer ownership model; Suno training liability disputed but less visible
• Winner: Music (liability more obscure)

Outcome by January 2026

• Tilly Norwood: Blacklisted; no agency representation; legally unusable on signatory productions
• Xania Monet: $3 million deal; charted on Billboard; 465k monthly Spotify listeners; cultural legitimacy

Strategic Implication: Producers will likely focus synthetic talent development on music, gaming, social media (TikTok), and advertising—where union protections are weakest—while avoiding theatrical/TV film where SAG-AFTRA enforcement remains formidable.

---

8. Intellectual Property Ambiguities: Who Owns Synthetic Performers?

A critical unresolved question is copyright ownership of synthetic performers.

Xania Monet’s IP Structure

• Lyrics/Creative Direction: Owned by Telisha Jones
• Voice/Audio Generation: Generated by Suno platform (ownership disputed)
• Visual Persona: Generated by Suno (ownership disputed)
• Record Deal: Hallwood Media signed contract with Jones; Jones assigned/licensed rights to label

Question: Does Suno retain ownership claim over the voice/visuals it generated? If Xania’s voice is sued for copyright infringement (for unauthorized training on copyrighted singers), who is legally liable—Jones or Suno?

Copyright Office Guidance (Incomplete)

The US Copyright Office has stated that:

• AI-generated works alone cannot be copyrighted (requires human authorship)​
• Films containing AI-generated material retain copyright protection if human-authored portions exist​
• AI content only copyrightable if human has determined “sufficient expressive elements”​

Gap: This guidance doesn’t clarify who owns AI-generated voice/likeness when:

1. The voice is used commercially (Xania Monet’s $3M deal)
2. The voice was trained on copyrighted source material (unknown corpus)
3. Multiple parties contributed (AI platform + human handler)

Training Data Liability (Unsettled)

• RIAA lawsuits against Suno/Udio allege copyright infringement for training on copyrighted music without permission​
• NO FAKES Act: Doesn’t address pre-creation training liability; only covers post-creation distribution
• California AB 2602: Doesn’t prohibit training AI on performer footage without consent

Result: Xania Monet’s voice may have been trained on copyrighted singers’ recordings without compensation—but neither Suno nor Jones has been held liable, and it’s unclear whether they could be.​

---

9. Comparative Legal Landscape (January 2026)

Synthetic Performer Legal Protection: Comparative International Landscape (Jan 2026) 

Four major jurisdictions are developing synthetic performer protections with divergent approaches:

California (Enacted)

• Status: AB 2602 & AB 1836 effective January 1, 2025
• Strength: Prohibits unfair contracts; protects deceased personalities
• Weakness: Vague “reasonably specific” standard; training data liability not addressed; may not preempt broader state laws; unclear preemption interaction

Federal (Pending)

• Status: NO FAKES Act in committee (revised version, January 2026)
• Strength: Creates federal property right in voice/likeness; DMCA-like safe harbor for platforms; subpoena power for rights holders
• Weakness: Post-expiration reuse loophole; 10-year limit undermined; ambiguous preemption language; training data liability unresolved; minors protection incomplete; burdens Copyright Office with registration

UK (Union-Negotiated)

• Status: Equity balloted for industrial action (December 2025); negotiations ongoing with Pact
• Strength: Union solidarity strong (90% voted for strike); future agreements could exceed US protections
• Weakness: No statutory Right of Publicity law; union protections only bind signatory producers; non-union productions unprotected

EU (Fragmented)

• Status: EU AI Act applies transparency requirements; member states have varied right of publicity laws
• Strength: Emerging consensus on deepfake harms
• Weakness: No harmonized digital replica law; patchwork approach enables regulatory arbitrage; training data liability unresolved

Strategic Gap: Producers can route synthetic performer production through jurisdictions with weakest protections (e.g., non-union UK, some EU states, non-signatory US studios).

---

10. Unresolved Questions & Future Scenarios

Copyright & Ownership

1. Who owns synthetic performer IP?
   • Creator of AI system (Suno, ElevenLabs)?
   • Human handler (Telisha Jones)?
   • Label/studio (Hallwood Media)?
   • Joint ownership with contested rights?
2. Can training on copyrighted performances proceed without authorization?
   • RIAA litigation pending; outcome will determine Xania Monet’s legal status
   • If training deemed infringement, does Xania face retroactive liability?
3. Are synthetic performers’ descendants entitled to postmortem publicity rights?
   • If Xania Monet “lives” 50+ years, can Telisha Jones’ estate license her after Jones dies?
   • Can a synthetic performer outlive its human creator?

Labor Market & Compensation

4. How will union pension funds remain solvent if synthetic performers displace 20-30% of background actors by 2030?
   • Current SAG-AFTRA pension funded by per-capita contributions; lower membership = insolvency
   • Proposal: Tax synthetic performer licensing fees to fund performer retraining?
5. Will “buyout culture” normalize, permanently ending residual income?
   • Current trend: 96% of performers report increasing buyout pressure
   • If synthetic alternative available, will residuals disappear entirely?

Enforcement & Regulatory Arbitrage

6. How will global enforcement work if synthetic performers created in low-protection jurisdictions?
   • NO FAKES Act (if passed) only covers US; international streaming platforms may route synthetic talent through non-signatory countries
   • Example: European producer creates synthetic actress, sells to US distributor; who has enforcement jurisdiction?
7. Will non-union productions become the default as studios avoid SAG-AFTRA synthetic performer premiums (1.5x scale)?
   • SAG-AFTRA enforcement depends on signatory status; non-signatories exempt
   • Incentive structure may push production away from union toward non-union indie model

---

11. Conclusion: The Bifurcated Future

Synthetic performers are not arriving as a sudden wholesale replacement of human talent. Instead, they are infiltrating the entertainment industry along two divergent paths:

Path 1: Film/TV (Union-Protected, Residuals Intact)

• SAG-AFTRA protections hold strong; Tilly Norwood rejected
• 1.5x scale premium makes synthetic casting economically unattractive
• Residual income model (weakened but intact) preserved
• Outcome: Slow, incremental synthetic adoption only where cost savings exceed union premium costs
• Risk: Medium-term (non-union film may emerge as workaround)

Path 2: Music/Influencing/Advertising (Union-Weak, Residuals Nonexistent)

• Xania Monet succeeds; weaker union enforcement
• Training data liability unresolved; corporate incentives favor synthetic
• Residuals already eliminated; synthetic adds minimal incremental harm
• Outcome: Rapid synthetic adoption; background music/influencing/commercial presenters displaced 2026-2028
• Risk: High (entire career path disrupted)

Long-Term Structural Risk: Even if union holds strong in film, the bifurcated market creates a vicious cycle—synthetic talent dominates in music/advertising, displacing music/voice actor income, pushing those performers into film, increasing supply and depressing film wages. The union loses scale (fewer profitable entry-level roles), reducing negotiating power in future contracts. Eventually, the economics of scale tip toward synthetics even in union productions.

The Residual Threat: The deepest economic threat to entertainment labor is not employment loss per se, but the permanent destruction of the residual income model. Once studio contracts normalize around buyouts (driven by synthetic competition), the long-tail income stream that funded pensions, healthcare, and middle-class actor stability will vanish. A profession that required $50,000/year in residuals to remain viable becomes impossible to sustain.

Policy Implications: Protecting human performers will require more than DMCA-style notice-and-takedown protections (NO FAKES Act). It will require:

1. Statutory residuals for synthetic performer licensing (i.e., synthetic use triggers mandatory payments to displaced performers)
2. Training data compensation (AI developers must license or compensate for performances used in training)
3. Synthetic performer registration (all synthetics must be disclosed and registered; no anonymity)
4. Non-union enforcement (protections must bind non-signatory producers, not just union signatories)
5. International harmonization (prevent regulatory arbitrage by routing synthetic production through weak-protection jurisdictions)

The Tilly Norwood incident was a victory for labor unions, but Xania Monet’s success suggests that victory was tactical, not strategic. The longer battle—over residuals, pension solvency, and the viability of entertainment as a profession for non-stars—is just beginning.

XYZ123

Comments are closed.