Protecting the Rights of AV Performers in the Age of Artificial Intelligence

As artificial intelligence (AI) transforms the entertainment landscape, it brings both innovative opportunities and significant risks, particularly for actors and other audiovisual performers. AI can replicate performers’ voices, images, and movements with unprecedented fidelity, yet such replication can undermine performers’ control over their identity and work. This paper advocates for ethical standards and protective measures that balance industry growth with respect for performer rights. We present a framework prioritizing informed consent, fair compensation, transparency, and legislative support.

AI’s capability to simulate human performances has led to new ways of creating content, from dubbing to digital doubles in movies. However, this rapid development raises concerns about creative ownership, privacy, and fair compensation. This policy paper addresses these concerns and underscores the necessity for a balanced approach that enables innovation while protecting the identities, rights, and incomes of AV performers.

Current Challenges:

• Identity Theft and Misuse: Unauthorized AI-generated content can misrepresent a performer’s values, beliefs, or artistic choices.
• Economic Displacement: AI could replace human performances, jeopardizing jobs, wages, and future opportunities for performers.
• Lack of Consent and Control: AI technology often bypasses traditional contractual and consent structures, leaving performers vulnerable to unfair practices.

By setting clear protections and industry standards, GAVA aims to advocate for ethical AI practices that benefit both creators and audiences.

1.Guiding Principles

GAVA proposes the following principles as a foundation for fair and ethical AI use:

a.Informed Consent: Performers must have the right to understand and authorize:

• The making and use of any AI-generated replica, ensuring complete control over their digital likeness Performers should also have final approval over any AI-generated content replicating their personal features.
• Any repurposing of the AI replica beyond the terms initially agreed, free form professional or economic retaliation.
• The use of their recorded or fixed performances to train AI models.

b.Fair Compensation: A robust compensation model should recognize the long-term value that AI-generated content can bring to producers, ensuring performers are justly rewarded.

c.Right to Opt-Out: If the use of recorded or fixed performances for training AI models is conducted under a statutory license or exception, performers must have an unwaivable right to effectively exclude all or some of their recorded of fixed performances from the scope of such license or exception.

d.Transparency and Accountability: All AI use should be publicly disclosed, and providers and deployers must be held accountable for unauthorized usage.

e.Data Privacy and Control: Rigorous privacy protections must govern the storage and use of a performer’s data, adhering to international privacy standards.

2.Policy Recommendations

A: Consent and Authorization Requirements

Performers should retain exclusive control over how their image, voice, and likeness are digitally replicated. This policy recommends the following:

• Explicit Written Consent: Performers must provide written and informed consent for any AI usage, specifying where, how and when it will be used, as well as any usage limitations. They should also have final approval over any AI-generated content replicating their personal features. Performers should have the right to object to any AI-replication of their image, voice and likeness in their contract, without facing any penalties or prejudice.
• Interpretation of contracts: The legal principle that consent and assignment of rights shall not cover methods, uses or means of dissemination that did not exist or were unknown at the time of the assignment must prevail in the interpretation of contracts.
• Time-Bound Consent: Consent should be restricted to specific timeframes and renewable only with the performer’s approval, preventing indefinite, unauthorized use of their image, voice and likeness.
• Defined Scope and Purpose: Any deviation from the scope of the original agreement must require a previous, informed and additional consent. This ensures performers may not be unknowingly represented in contexts that might harm their reputation or career.

B:Fair Compensation for AI-Generated Performances

As AI-generated performances become a lucrative aspect of the media landscape, fair compensation for performers is essential. Here are recommended provisions:

• Initial Compensation and Ongoing Remuneration: Performers should be paid when their image, voice and likeness is digitally replicated or enhanced, and receive additional and appropriate remuneration payments for each future use or distribution of the resulting content previously expressly agreed to.
• AI Training Data: Any recorded or fixed performance used to train AI models must be treated as part of the performer’s intellectual property. Performers should be compensated for this data, recognizing it as a core part of the AI’s value.
• Revenue Sharing Model: For large-scale uses of AI, performers should participate in a revenue-sharing model, ensuring they benefit proportionally from the success of AI-enhanced projects.

C:Transparency and Disclosure Requirements

To maintain public trust, it is crucial that AI usage in content creation is transparent and accountable. Key recommendations include:

• Clear AI Use Disclosure: Productions must clearly disclose AI usage, both to performers and to audiences. This could be accomplished via end credits, disclaimers, or adding accessible and visible information on online databases.
• Detailed Contracts: Contracts shall include a reasonably specific description of the intended use of any AI-generated replica, with a clear, accessible breakdown of all potential applications, duration, and revenue structures. 
• Third-Party Audits: Regular third-party audits can ensure AI is deployed in compliance with ethical standards and contractual agreements.

3.Advocating for Legislative Support

GAVA strongly advocates for international and national legislation that addresses the rights and protections of performers in the AI era. We recommend:

• Strengthening Copyright and Personality Rights: Performers’ performances, images, voices, and likenesses should receive legal recognition as protected forms of their intellectual and personal property. Personality rights protections should allow performers to prevent AI usage that misrepresents them.
• AI-Specific Privacy and Usage Laws: Governments should establish legal frameworks restricting unauthorized data usage for AI, particularly for biometric data. These laws would mirror privacy-focused initiatives like the GDPR in Europe.
• Industry Standards for AI Usage: Industry coalitions should be encouraged to establish AI codes of conduct, collective agreements and/or model contracts with enforceable guidelines for AI usage. These standards can create a foundation for industry-wide accountability.

Conclusion

The swift advancement of AI technologies requires a clear policy framework that prioritizes performers’ rights without stifling innovation. This paper recommends an approach rooted in informed consent, usage limitations and fair compensation, alongside industry transparency and a strong policy framework, acknowledging economic and moral rights to performers, including with respect to AI-use, and protecting their personal data, including biometric data, from unauthorized exploitation.

Given the rapid evolution of AI and its applications, it is essential for any regulation in this area to remain open to periodic reviews. This will allow stakeholders to address emerging challenges, adapt to technological advancements, and ensure continued protection for performers in an ever-changing landscape.

GAVA calls on industry stakeholders to adopt these standards, ensuring a sustainable and respectful approach to AI in the entertainment industry.