The AI brand: the boundary of authenticity.
AI can assist with many tasks, but it cannot replace real experience, and using it incorrectly can erode trust faster than it can be built.
The unchanging core principle: all brand statements must be truthful and substantiated; using AI does not exempt you from responsibility for what you say. When AI creates or significantly edits external content, especially advertisements and testimonials, it should be clearly disclosed to customers. Concealing information at the right time must be transparent, as it is the quickest way to lose trust when discovered.
- AI only supports backend drafting and organizing
- Clearly define the role of AI with customers.
- content that does not affect the authenticity of claims
- fabricating testimonials or reviews using AI to appear authentic
- impersonating voice or likeness of real people without permission
- conceal the role of AI where transparency is expected
AI tools can now create testimonials that look like they are from real customers, build virtual influencers with extensive activity histories, and mimic the voice or image of real people in just a few minutes. That technique is not inherently bad, but when brands use it to create a false impression of reality, the ethical boundaries are crossed, and the legal risks that follow are significant.
Why authenticity is an asset
Trust is something a brand accumulates over many years and can be lost in a few days if customers discover they have been misled. In the age of AI, the speed at which a "leak" spreads has significantly increased due to the growing prevalence of tools that detect AI-generated content and the rapid response of online communities. Brands built on authenticity stand stronger when the environment changes. Brands based on illusion are more likely to collapse when the illusion fades.
The boundary must not be crossed.
There are some clear boundaries that do not depend on brand perspective or style.
- Fake testimonials and reviews: AI should not be used to fabricate praise, product reviews, or customer stories in a way that suggests they are real consumer experiences when they are not. This is deceptive advertising, not just an ethical violation.
- Imitating the voice or image of real people: Whether a celebrity or an ordinary person, imitating their voice, image, or style for commercial purposes requires both consent and clear disclosure to the audience. Lacking either is wrong.
- False claims covered by AI: "AI wrote it, so it's not my fault" is not a valid argument. All claims in brand documents, whether by AI or a human, must be truthful and substantiated.
When to need transparency, when not needed
You should disclose the use of AI: Advertising content, testimonials, representative images, product introduction videos. These are elements that customers are led to believe in a specific experience or person.
Do not necessarily label: AI assists behind the scenes, such as drafting content, organizing ideas, and optimizing manuscripts. When the tool only helps the writer work faster without changing the essence of the statement, there is no need to annotate AI everywhere. Excessive labeling creates noise and does not increase trust.
The boundary lies in the question: What are customers being led to believe, and is that true? If they believe this is a genuine experience of real people when it is not, transparency is needed, regardless of whether AI is involved.
Common errors in practice
- Using AI to create user profiles or customer stories and then publishing them as real case studies.
- Asking AI to compile multiple real feedback into a "representative" and present it as if it were from a specific person.
- Creating a virtual influencer without disclosure, allowing the audience to perceive them as a real person.
- Hiding the use of AI in places where it should be mentioned, and then losing trust when it is revealed, is much worse than being upfront about it.
- Conversely, labeling everything with AI, even where it’s unnecessary, creates a sense of brand insecurity or inconsistency.
All statements must be truthful and substantiated. Using AI does not excuse misleading statements.
FTC staff guidance on AI in advertising, 3/2025
The viewpoint of Sinh Vũ
Sinh Vũ prioritizes authenticity over convenience. In practice, this means: no creating fictional clients, no imitating real people, and no letting AI speak for an experience that doesn't exist. AI supports craftsmanship, not impersonation.
For Sinh Vũ's clients, the boundary is set from the strategy phase: every statement in the brand documents must hold up to scrutiny if checked. If you are unsure whether a specific piece of content needs to be noted, the simplest check question is: If customers knew this, would they feel deceived? If yes, then transparency is needed. If not, then there’s no need to worry.
Transparency in the right places does not weaken the brand. On the contrary, it shows that the brand is confident enough not to need illusions.
Decision checklist
Topic: Ethics and authenticity of brands in the age of AI. Sinh Vũ guide, sinhvu.com
Select each item you find appropriate, then print or save as PDF to take with you.
If you have marked most of the signs above, this is the time to discuss in more detail. Sinh Vũ can help you review and propose a direction.
References
FTC AI Content Disclosure Rules (ppl.studio, summary guidance 2025-2026); FTC staff guidance on AI in advertising (3/2025); FTC guidance on AI likenesses (2025). The practical commentary is the experience of Sinh Vũ Studio.
Frequently asked questions
Is it necessary to clearly state that AI was used to write social media posts?
It's not always necessary. If AI only supports backend tasks like drafting and organizing ideas, while the content declaration is genuine and originates from the brand, labeling is not mandatory. The boundary lies in whether the declaration is accurate, not who typed it.
Do former clients allow their quotes to be used, and can Sinh Vũ ask AI to rewrite them more smoothly?
Acceptable, as long as the meaning and real experience remain intact, without adding anything they did not say. If AI significantly edits to the point that the testimonial changes content, it should be confirmed with that person before use.
I want to create a brand representative character using AI. Is that possible?
Yes, but the character must be clearly fictional and should not imply that this is a real person or a real customer experience. If the character mimics the appearance or voice of a real person, permission must be obtained from them and disclosed to the audience, both conditions must be met, not just one.