The New Rules for Representing Public Figures Through AI

Generative AI is changing what it means to represent a person after they have left the stage, playing field, classroom or public eye.

A voice can be recreated. A likeness can become interactive. A historical personality, sporting icon or entertainer can appear as a Digital Human capable of answering questions in real time.

This creates possibilities that extend far beyond conventional photographs, recordings and archive footage. It also introduces an important question:

Who should control the authorised AI representation of a public figure?

The answer becomes more complicated when an individual’s identity carries significant cultural or commercial value, or when that person is no longer able to approve how their likeness, voice and personality are used.

Ailias is developing an authorisation-first model for conversational AI Digital Humans, designed to be created in partnership with rights holders, estates, foundations, talent representatives and other responsible organisations.

This reflects a broader shift in the conversation around generative AI: from asking what the technology can create to determining who has the authority to approve, guide and govern what is created.

From Digital Replica to Conversational Identity

Traditional digital representations reproduce what already exists, photos capture appearance, recordings preserve voice, and film documents words or actions.

Conversational AI goes further, allowing a Digital Human to generate new dialogue in real time using a recognisable appearance and voice. This raises important questions about authorisation, information, personality, approved topics, changes, and who can suspend or retire the representation. Clear governance helps establish the conditions for audience trust.

Permission Is the Beginning

Recreating someone’s appearance or voice does not automatically grant permission to present it as an official representation.

Living figures may require authorisation from the individual, representatives, or licensing partners. For historical figures, rights may involve estates, families, foundations, museums, or other rights holders.

Different elements, such as name, likeness, voice, archives, trademarks, and written works, may have separate owners. Rights verification is therefore essential.

But permission to create a representation is only the starting point. A conversational identity also requires ongoing decisions about its knowledge, behaviour, personality, and control.

From Authorisation to Stewardship

Initial permission determines whether a representation may be created and how it may be used. A conversational representation also requires continuing decisions about how it behaves after launch.

Rights-holder permissions therefore need to be translated into practical controls. These may include:

• Approved knowledge sources
• Personality and behavioural guidance
• Restricted or prohibited subjects
• Commercial and brand-association boundaries
• Procedures for testing and approval
• Monitoring and audit records
• Change-control responsibilities
• Processes for correction, suspension and retirement

This is the distinction between licensing a likeness and stewarding an interactive identity.

The first grants permission for a defined use. The second governs how that identity continues to represent the individual throughout its operational life.

Ailias describes this as stewardship because the responsibility does not end when the Digital Human is launched. It continues for as long as the representation remains available to audiences.

Authenticity Is More Than Looking Real

A Digital Human may look and sound remarkably similar to the person it represents. Visual and vocal accuracy can make an experience compelling, but resemblance alone does not establish authenticity.

For conversational AI, authenticity also depends upon the sources behind its answers, the way its personality has been defined, the boundaries governing its behaviour and the authority of the organisation responsible for it.

An unauthorised imitation may appear highly realistic. An authorised representation may deliberately avoid answering certain questions because those boundaries protect the individual, the rights holder and the audience.

Authenticity is therefore not simply a measure of visual fidelity. It also involves provenance, permission and accountability.

Audiences should be able to understand that they are interacting with an AI representation rather than the person themselves. They should also be able to identify the organisation responsible for authorising and operating that experience.

Living Figures and Evolving Identities

Governance is particularly important when representing a living person.

A sporting figure’s career can change quickly. An entertainer may enter new commercial relationships. A business leader may change roles. Public views, personal circumstances and professional priorities can evolve.

An authorised representation cannot simply be created once and assumed to remain accurate indefinitely.

Its knowledge may need updating. Previous answers may become outdated. New commercial restrictions may arise. Subjects that were once appropriate may later become sensitive. The individual may also wish to change how their identity, personality or current position is represented.

This makes change control an important part of the relationship. Someone must have the authority to approve updates and decide whether the Digital Human should continue operating in its existing form.

Where possible, creating a representation with the living individual also provides an opportunity to record their intentions directly: what they would welcome, what they would reject and how they would want their identity to be treated in the future.

Protecting Historical and Cultural Legacies

Estates, foundations, museums and cultural organisations face a different but equally important responsibility.

A conversational historical figure can make knowledge more accessible, support multiple languages and encourage audiences to explore a life or subject through their own questions. Used thoughtfully, this can create an unusually personal form of education and interpretation.

However, historical personalities cannot correct an inaccurate response, object to an inappropriate association or explain how they would wish to be represented.

The organisations responsible for their legacy must therefore consider whether a conversational representation is appropriate at all, and, if it is, how historical evidence, uncertainty and disputed interpretations should be handled.

Sometimes the responsible decision may be not to create a Digital Human of the central figure. A fictionalised or composite character may offer a more appropriate way to interpret the period without attributing newly generated words to a real individual.

Responsible innovation includes knowing when the technology should not be used.

Creating Legitimate New Opportunities

Handled carefully, officially authorised Digital Humans could create valuable new opportunities for public figures and rights holders.

Potential applications include:

• Museums, exhibitions and visitor attractions
• Sporting venues and fan experiences
• Multilingual educational interpretation
• Brand partnerships and licensed activations
• Hospitality and premium experiences
• Archive-led storytelling
• Online fan engagement
• Masterclasses and knowledge-sharing
• Authorised legacy and remembrance experiences

A representation could be developed as a governed identity and then used across individually approved applications rather than recreated independently for every project.

This may create new licensing and commercial opportunities while allowing the rights holder to retain meaningful control over where the identity appears, what it may discuss and how it is presented.

The value does not come simply from recreating a recognisable face. It comes from creating an experience that audiences want to engage with and that the represented individual or responsible rights holder is prepared to stand behind.

Governance Does Not End at Launch

Conversational AI, information sources, and audience expectations evolve over time. New questions, risks, technology changes, or knowledge sources can affect how a Digital Human behaves.

Ongoing stewardship should include reviewing conversations, maintaining approved sources, testing updates, and keeping the representation within its agreed purpose.

Clear responsibility is equally important: Who can change its knowledge or behaviour? Who approves updates? Who handles challenged answers? Who can suspend it? What happens when a licence or responsible party changes?

A strong governance model connects rights holders, producers, technology providers, and operators through clear permissions and approval processes.

Who Controls the Authorised Representation?

A Digital Human may combine a person’s name, likeness, voice, biography, personality, archives, and intellectual property with AI and presentation technology.

No single supplier should assume that creating one component gives them ownership of the represented identity. Likewise, owning the identity or archives does not determine ownership of the technology or production assets.

Agreements should clearly define each party’s rights and responsibilities, while keeping ownership of the person’s identity separate from ownership of the technology representing it.

Resources for Rights Holders

Ailias has developed resources to help rights holders and organisations begin considering these questions.

The Digital Replica Rights Guide examines permission, likeness, voice, licensing and the practical issues rights holders should consider before approving an AI representation.

The Ailias AI Stewardship Framework addresses the continuing governance of conversational Digital Humans, including knowledge, behaviour, approval, monitoring, change control and accountability.

Together, these resources reflect a central principle: having the technical capability to reproduce aspects of a person’s identity is different from having the authority and governance required to operate an official representation.

A New Responsibility for the AI Era

A photograph reproduces an image; a recording preserves a moment. Conversational AI can create an entirely new interaction.

That makes permission essential, but not sufficient. Rights holders and technology partners must also define what the representation knows, how it behaves, who controls it, and who remains responsible.

As digital representations become more common, the key question is not simply whether someone can be recreated, but whether that representation is authorised, authentic, and responsibly managed.

Learn more about Ailias and its approach to responsible AI representation, or connect with the company on LinkedIn, YouTube, Instagram, and TikTok.