Rakuten Research Examines AI Disclosure and Trust in Creator Reviews

Only 38% of surveyed consumers would be comfortable with an entirely AI-generated creator review, according to Rakuten Advertising’s November 2 research update. Acceptance was higher for translation and editing.

The findings arrive as brands prepare holiday campaigns involving product recommendations and shopping content. For affiliate managers, they raise questions about which production tasks creators can delegate and how audiences understand the use of AI.

Key Takeaways From the Rakuten Creator Trust Research

• 54% were comfortable with AI translation, compared with 46% for photo or video editing.
• AI disclosure would increase trust for 45%, reduce it for 26%, and make no difference for 29%.
• Responses differed across countries and age groups.
• The findings describe stated attitudes and do not establish a change in sales.

AI Production Tasks and Product Opinions

An AI tool can support several stages of a creator’s work. Translating a caption after a product test, shortening footage, and generating an assessment involve different decisions about what the audience sees and whose judgment supports it.

For example, a creator reviewing a bag could use software to translate their description of its compartments. Readers could still assess the original demonstration. If software supplies an opinion about comfort or durability, the reviewer would need evidence for that claim before presenting it as personal experience.

For affiliate publishers, documenting the basis of a recommendation helps readers distinguish manufacturer information from independent testing. An AI disclosure explains part of the production process, but it cannot demonstrate whether the reviewer used the product.

Separate IAB research, discussed in Affiverse’s coverage of creator reviews and trust in AI shopping, found that 65% of recent AI-shopping users felt more confident when credible creator reviews informed recommendations. That study surveyed 2,200 people across five markets. It examined creator input within AI shopping advice, a different question from how audiences react when creators generate reviews with AI.

AI Disclosure and Audience Differences

Rakuten reports that disclosure produced the strongest stated trust gains in Brazil and the US, while Australia was nearly evenly split. Millennials were the most receptive age group, with 52% saying disclosure would increase trust.

Those variations give campaign teams a reason to examine audience responses before assuming that one explanation will work equally well everywhere. A short statement about translation describes a different intervention from a notice that product imagery or recommendations were generated.

Commercial relationships also need their own explanation. Affiverse’s reporting on Meta’s AI labels for Facebook and Instagram ads explores the distinction between identifying AI involvement and identifying who benefits from a promotion. An AI label does not, by itself, tell a shopper whether a creator receives commission or has been paid by a brand.

What the Rakuten Survey Measures

Rakuten attributes the findings to its September 2026 Holiday Shopping Intentions survey with Suzy across the US, UK, Australia, France, and Brazil. The summary does not give the sample size, weighting details, or questionnaire.

The results describe how people say they would respond to AI use and disclosure. Comfort, trust, clicking a link, and completing a purchase are separate outcomes. A respondent expressing greater trust has not necessarily made a purchase, and these percentages cannot be used as a forecast for affiliate conversion rates.

Brands assessing commercial performance would need their own campaign evidence, including attributable orders and validated sales. Comparing campaigns would also require accounting for differences in products, offers, and audiences.

Affiverse Take on Holiday Creator Briefs

Affiliate teams can use the findings as a prompt to make creator agreements more specific about production and review standards. A brief could identify permitted AI uses, explain which product claims require supporting evidence, and assign responsibility for checking the finished content.

For a product demonstration, that could mean retaining original footage and confirming that edits preserve the conditions shown during testing. For a translated review, a language check could confirm that limitations and qualifications remain intact. These are practical editorial checks; the survey does not measure their effect on revenue.

Affiverse’s Black Friday affiliate marketing strategy places creator preparation alongside confirmed offers, commercial terms, and tracking tests. Adding production and disclosure expectations to that process would give partners clearer instructions before content enters approval, while leaving sales performance to be assessed through campaign reporting.