AI affects brand trust through the choices customers can see or later discover: whether imagery shows the real product, whether synthetic content is disclosed, whether claims are true, and whether a person is accountable for what is published. Trust is rarely lost because a business used AI. It is lost when AI was used to mislead, or when nobody checked the output.
Where AI decisions become trust decisions
| Decision | The trust risk | A safer default |
|---|---|---|
| Product imagery and video | Customers expect what they see | Show the product as it really is, using real references |
| Realistic people, voices or events | Viewers may assume footage is real | Disclose synthetic content where it could mislead |
| Copy and claims | Fluent text can state things that are wrong | Substantiate every factual claim before publishing |
| Customer-facing automation | People may not know they are dealing with software | Say so, and offer a route to a person |
| Customer data in AI tools | Privacy expectations | Do not paste personal information into tools you have not assessed |
Accuracy comes first, and the law agrees
The ACCC’s guidance on misleading claims states that businesses must not mislead consumers, that intent does not matter, and that claims in ads, packaging, social media and websites must be accurate and have reasonable grounds. It also notes that leaving out important information can mislead. For AI-generated work this has a direct meaning: a generated product shot that exaggerates a finish or feature, or an AI-written testimonial presented as a customer’s, creates the same exposure as any other misleading content. This is general information, not legal advice, so seek professional advice for specific campaigns. Related reading: product photography versus AI-generated product imagery.
Disclosure: follow the platform and the audience
Platforms are setting their own rules. YouTube’s help page on altered or synthetic content asks creators to disclose realistic content that is generated or meaningfully altered with AI, for example a real person appearing to say or do something they did not. It lists exemptions, including clearly unrealistic content and AI used only for production help such as scripts or brainstorming, and says labels can be applied automatically, including when C2PA metadata is detected. Disclosure on its own does not limit a video’s audience or monetisation eligibility, according to that page. Rules differ by platform and change, so read the current help page for each channel before publishing.
Provenance technology is also emerging. The C2PA explainer describes Content Credentials as signed records of an asset’s origin and edits, including use of AI. It is candid about limits: validation shows an asset has not been tampered with since the credential was created, not that its content is true, metadata can be removed, and missing credentials do not prove content is untrustworthy.
A trust-centred AI policy outline
A one-page policy is enough for most small and mid-sized businesses. Cover:
- Purpose: where AI may be used (drafting, ideation, production) and where it may not (for example, fabricating reviews or faces of real customers).
- Accountability: a named person approves anything public-facing.
- Substantiation: facts, prices and performance claims are checked against source material.
- Disclosure: when and how synthetic content is labelled on each channel.
- Real-world fidelity: products, results and premises shown must match reality.
- Data handling: what information may and may not enter AI tools.
- Review: revisit the policy at least every six months, since platform rules shift.
Honest limitations
Customers differ on how much they care. Some audiences are relaxed about AI-assisted creative work and others are sceptical, so test reactions rather than assuming. Disclosure can also be clumsy: a label that reads like an apology can hurt more than quiet, accurate use. The aim is to be unsurprising: nothing a customer discovers later should feel hidden.
See also the real limitations of AI-generated video and AI video production for brands for how these choices play out in production.
Where Radiance Impact fits
Radiance Impact combines creative production, AI, digital technology and marketing, and treats these questions as part of the work. Learn more about the studio or contact us to talk through an AI policy for your brand.
Sources
- False or misleading claims, ACCC
- Disclosing use of altered or synthetic content, YouTube Help
- C2PA and Content Credentials Explainer, Coalition for Content Provenance and Authenticity
