The Short Answer Your Legal Team Will Appreciate
AI-generated actors and synthetic avatars can be used effectively in commercial video. But before you brief a production company or approve a storyboard featuring a digital human, there are consent, intellectual property, and reputational questions that need clear answers. If you are a brand marketer or compliance officer in Singapore, this article gives you a practical framework for making that decision responsibly.
Synthetic talent is no longer a novelty. Digital humans now appear in product explainers, banking communications, retail campaigns, and regional brand films across Southeast Asia. The technology has matured, but the governance around it has not always kept pace.
What "Synthetic Talent" Actually Means in a Production Context
The term covers a wider range than most briefs acknowledge. Before approving any production approach, it helps to distinguish between the main categories:
Fully original AI-generated characters
These are digital humans built from scratch using generative tools. They are not modelled on any specific real person. The character's face, voice, and mannerisms are synthesised rather than derived from a human performer. From a consent standpoint, this is the cleanest category, but it still carries brand and audience-perception considerations.
AI-cloned or digitally replicated likenesses
This involves training a model on the image, voice, or movement data of a real person, whether a professional actor, a brand spokesperson, or in some cases a public figure. This is where consent and contractual obligations become critical. Using a person's likeness without documented, specific, and informed consent is a serious legal exposure, regardless of how technically convincing the output is.
AI-assisted performance enhancement
Some productions use a real human performer on set and then apply AI-driven visual effects in post-production: de-ageing, digital doubles for stunt sequences, or voice synthesis to correct a line reading. This sits in a middle category. A real performer is involved, but the final output may diverge significantly from what they actually delivered.
Licensed synthetic talent
A growing number of talent agencies and AI platforms offer synthetic actors whose likenesses have been built with the performer's consent and are available for commercial licensing. These operate similarly to stock photography: you pay for a usage tier, and the consent chain is managed by the platform. This is often the most practical starting point for brands without an existing spokesperson relationship.
The Consent and Legal Checklist for Singapore Brands
Singapore does not yet have a single statute that governs AI-generated likenesses specifically, but several existing legal frameworks apply and should be reviewed with qualified legal counsel before production begins.
Personal Data Protection Act (PDPA). If the synthetic avatar is derived from biometric or image data of a real, identifiable individual, data protection obligations apply to how that data was collected, stored, and used.
Common law passing off and defamation. Creating a digital likeness of a real person and placing them in a commercial context without consent can give rise to passing-off or reputational harm claims, particularly if the content implies endorsement.
Contractual obligations with existing talent. If your brand has an existing contract with a human spokesperson or ambassador, check whether that agreement restricts the use of AI-generated likenesses that may substitute for or compete with their contracted appearances.
Intellectual property ownership of the output. Clarify upfront who owns the synthetic character. If your production company created the avatar using a third-party AI platform, ownership may rest with the platform, the studio, or your brand depending on the licensing terms. Get this in writing before the asset is built.
Disclosure to audiences. While Singapore does not currently mandate explicit disclosure that a video features synthetic talent, there is a reasonable expectation of transparency in advertising. The Advertising Standards Authority of Singapore (ASAS) guidelines on truthful representation are relevant, and in regulated industries such as financial services and healthcare, additional rules may apply.
Practical Production Decisions That Reduce Risk
The legal and ethical questions are not reasons to avoid synthetic talent. They are a checklist to work through before production, not after. Here is how responsible production teams approach it.
Build the consent chain before building the asset. If your brief calls for a specific real person's likeness, secure written consent that specifies the usage: which markets, which platforms, for how long, and for what category of product. Vague verbal agreements are not sufficient.
Use purpose-built licensed avatars where possible. For campaigns where brand recognition of a specific face is not the primary goal, licensed synthetic characters from reputable platforms offer a lower-risk path. The consent and IP chain is already documented.
Define the character's brief with the same rigour as a casting brief. A synthetic actor still needs to reflect your brand's values, demographic targeting, and cultural context. A financial brand in Singapore using a synthetic spokesperson should consider how that character reads across Chinese, Malay, and Indian audience segments. Generic does not mean neutral.
Retain production documentation. Keep records of the consent agreements, platform licensing terms, and the specific AI tools and model versions used. If a question arises after the campaign launches, you need an audit trail.
Brief your production partner on the compliance requirements upfront. A capable AI video production company will have frameworks for handling synthetic talent, but they need to know your industry's specific obligations before they begin creative development. Legal constraints should be in the original brief, not raised in the review stage.
When Synthetic Talent Is Probably Not the Right Call
There are circumstances where the practical and reputational risks outweigh the production benefits:
- Highly regulated product categories such as pharmaceutical or financial products may face stricter disclosure obligations or sector-specific advertising codes that make synthetic talent a compliance headache without sufficient benefit.
- Campaigns requiring emotional authenticity where the audience expects a real human connection, such as charity appeals or sensitive public health messaging, may find that a synthetic performer undermines rather than supports the message.
- Situations where audience trust is already fragile. If your brand is managing a reputational issue, introducing a synthetic spokesperson may raise further questions about transparency.
- Markets or audience segments unfamiliar with synthetic media. Not every audience in Southeast Asia has the same exposure to AI-generated content. Audiences who do not recognise the format may feel misled when they learn the spokesperson was not real.
For guidance on how AI-assisted approaches compare to traditional production across different campaign types, the article on AI video vs traditional production covers the broader trade-offs in practical terms.
What Good Production Governance Looks Like
The brands navigating synthetic talent most confidently are not the ones with the most advanced technology stack. They are the ones with clear internal governance: a documented approval process that involves both creative and legal review, a consistent policy on consent and disclosure, and a production partner who understands those requirements before the brief is written.
If your organisation does not yet have a policy on AI-generated likenesses, now is a practical time to draft one. The volume of synthetic talent in commercial production is increasing, and having a position in advance of a campaign brief is easier than building policy under deadline pressure.
For brands considering how this fits into a broader content production approach, the article on AI video production budgets for Singapore brands is a useful companion read.
Frequently Asked Questions
Do I need to disclose to viewers that a video uses a synthetic actor? Singapore does not currently have a law requiring explicit disclosure for all synthetic talent, but industry standards around truthful advertising apply. In regulated sectors such as finance or healthcare, additional rules may require transparency about who or what is endorsing a product. When in doubt, consult legal counsel and consider whether voluntary disclosure protects your brand's trust more than it costs.
Can I use an AI-generated voice that sounds like a real celebrity or public figure? Generally no, not without explicit consent. Using a synthesised voice that is recognisably modelled on a specific person, even without using their actual recordings, can give rise to passing-off and reputation claims. Licensed voice platforms that work directly with performers to offer sanctioned synthetic voices are a safer path.
Who owns the AI-generated character my production company creates for my campaign? This depends on the contracts in place between your brand, the production company, and the AI platform used to create the asset. Ownership is not automatic. It must be defined in writing before production begins, covering usage rights, exclusivity, the right to modify the character, and what happens to the asset after the campaign ends.
If your brand is planning a campaign that involves synthetic talent or digital human characters and you want a production partner who will address the consent and governance questions from the start, discuss your production brief with the team at Glory Forest Media.