[wadiz] Transparency Report Vol. 23

Issue date

2026. 10. 26

Target period

10.11 — 10.24

publication cycle

biweekly

Source: Fair Trade Commission, 「Artificial Intelligence (AI)-related labeling and advertising guidelines (draft)" on October 22, 2026.

The Fair Trade Commission has released a draft of AI-related labeling and advertising guidelines (listening to opinions by November 11). In the Korea Consumer Agency survey, 57.9% of respondents said they would buy AI-applied products even if they were more expensive, so the expression "AI" greatly influences purchasing decisions.

The point is that you can use the 'AI' expression, but you have to inform the actual application, the operating conditions, and the limitations of the performance. Simple sensors or timers that move according to the set rules are difficult to see with AI, and AI-made reviews should not be posted like actual reviews.  
Contents recommended by AI virtual people need to be marked with a 'virtual person', and food, cosmetics, and medicines will be banned from November 27th from the recommendation of AI-made virtual experts. 

Source: Ministry of Food and Drug Safety, 2026.10.16.

In collaboration with local governments, the Ministry of Food and Drug Safety will focus on medical institutions with high prescriptions for ADHD drugs from October 19 to November 18. Social media and online communities are also strengthening monitoring of illegal posts promoted and sold as "concentration-enhancing drugs."

During the CSAT period, food advertisements have also been repeatedly inspected. At the same time last year, 45 cases of unfair advertisements for food and health functional foods were detected during a special inspection, of which 29 (64.4%) were false and exaggerated advertisements with unrecognized functionality such as "improving memory."

Source: Korea Consumer Agency, October 2026

As a result of the Korea Consumer Agency's survey of passenger transport terms and conditions of major domestic and foreign airlines, a number of unfair provisions such as flight cancellation and refund restrictions, exemption from liability for airline negligence, and unilateral change of flight schedule were identified, and self-correction was recommended. Although the industry is different, phrases with the same structure are often seen on the funding detail page.

We recently identified a project to sell luxury scarves as a purchasing agent. The trust team looked at the official import and purchase routes with import declaration certificates and invoices, and requested additional documents to prove the authenticity, such as genuine product certificates. However, because the maker's calling was not sufficient and did not meet the terms of the funding process, Wadiz canceled the project.

The most important thing is whether the purchase agency reward is genuine or not. We first check the import declaration certificate and invoice, and if necessary, we can request the submission of a genuine product appraisal. If you prepare it in advance, you can submit it as soon as you request it.

[Case 2] "Reports of similarity, Wadiz sees it like this" — Criteria for identifying similarities

In the last issue, I introduced a case where a report was received, but the license was confirmed and proceeded normally, but the results were different this time. Recently, several reward projects have received reports of imitating famous brands or overseas designs, and the trust team has asked the maker for clarification on how it was made and its similarity to the original. As a result, projects whose similarities were recognized to be confused with the original were treated with no exposure.

Yes. Wadiz will continue to review the project through reporting and monitoring even after the project is opened, and if the problem is identified, we will take measures such as requesting clarification, requesting correction, and non-exposure. If you find anything suspicious about the open project, please register as [Report] on the project page.

This week, we pointed out safety transaction information such as AI labeling and advertising standards, health expressions for test takers, and non-refundable phrases, and looked at cases of luxury purchase agencies and design theft reports. When promoting 'AI', check the actual application range and operating conditions, and check before disclosure to see if there are any functional expressions that are not recognized for general foods.

Purchasing agency products were checked for official routes and authenticity with import declaration certificates, invoices, and genuine evidence, and design theft reports were non-exposed to projects that were recognized for similarity after comparative review with clarification.

Wadiz continues to check through reporting and monitoring even after the project is open. The maker should prepare proof of genuine product and license in advance, and the supporter should let me know if there is anything suspicious.

Shortcut →