In March 2026, the Global Research and Policy Institute (GRPI) submitted recommendations to the Cabinet Office’s Regulatory Reform Promotion Office, on measures to restrain advertising and labeling expressions that impede consumers’ voluntary and rational choice.
These recommendations were in response to a request for information (RFI) issued by the Regulatory Reform Promotion Office on regulatory and institutional barriers in the social implementation of artificial intelligence (AI). The RFI can be found here (in Japanese):
Details of the GRPI recommendations are as follows.
Background: development of AI tools for pre-screening of advertisements and representations
The Act Against Unjustifiable Premiums and Misleading Representations (hereafter referred to as the Act) prohibits businesses from falsely portraying the content and quality of goods or services as being significantly superior to actual goods or services, or to equivalent goods or services of competitors, and imposes surcharges or other administrative penalties for violations, to be issued by the Consumer Affairs Agency (CAA) (in serious cases criminal penalties may be imposed). However, reports of alleged violations of the Act submitted to the CAA have almost doubled to over 20,000 annually in the five-year period from FY2020. During this time, there has been almost no change in the personnel and budget funding devoted to implementation of the Act and consequently, the number of cases where administrative action has been implemented has hovered at around 30 to 40 per year.
Therefore, businesses need to strengthen their own initiatives to control misleading advertisements that obstruct voluntary and rational consumer choice. In recent years, some business operators have already introduced AI tools that can automatically sift through their huge portfolios of advertisements and representations to identify those that could potentially infringe the Act. The spread and implementation of such tools is expected to strengthen internal company pre-screening checks while keeping costs low and to reduce the number of alleged violations reported to the CAA.
CAA should publish grounds for judgment in all cases of administrative action pertaining to violation of the Premiums and Representations Act
Pre-screening AI tools need to improve their accuracy by sufficiently learning from cases of past violations in order to determine what were considered to be infringements of the Act.
Examination of published CAA documents regarding cases of administrative action taken over violations of the Act, shows that the authorities have included the grounds for judging such cases to be violations of the Act. However, in recent years, cases falling under “Regulations on advertising content that has not been substantiated” make up nearly 60% of cases of CAA administrative action. Under this regulatory system, if the business operator does not submit documents that objectively demonstrate the efficacy and effectiveness of the advertisement/representation within a set time period, the CAA will treat the advertisement/representation as misleading and issue an administrative action. In all such cases, the CAA’s published document merely states that the grounds for taking administrative action were that “Documents [submitted by the business operator] were not found to provide rational grounds to substantiate the representation.” Since no specific reason is included, AI cannot be trained to learn from such cases.
We believe that the CAA should publicize the grounds for judgment in all cases of administrative action pertaining to violations of the Act, in order that business operators can effectively control potential violations via their own pre-screening checks.
Strengthening powers of organizations operating voluntary industry rules to allow filing of injunctions for misleading advertisements
The authorities face a situation where administration actions cannot keep pace with increasing reports of alleged violations of the Act. We believe that, in addition to business operators strengthening their own pre-screening checks of advertisements and representations, another effective method to tackle this situation would be to allow other private-sector operators who have noticed misleading advertisements to file for injunctions against them.
Currently, Japan operates a framework where industry bodies (in most cases those with a name such as the XX Fair Trade Council) can prevent misleading advertisements based on the provisions of Fair Competition Codes, sets of voluntary industry-specific rules approved by the Japanese government. Such industry bodies can issue warnings and impose penalty fees when members violate the requisite Fair Competition Code. However, they do not have any direct control over misleading advertisements made by outsiders who are not a member of their organization. On the other hand, consumer organizations who fulfill certain criteria (known as “qualified consumer organizations” or QCOs) have the power to file for injunctions with the court against misleading advertisements, but there have been very few cases to date in which QCOs have actually exercised this power.
If the Japanese authorities were to grant Fair Trade Councils that fulfil certain conditions and are familiar with industry advertisements and representations due to their regular activities the same right to file for injunctions as is possessed by QCOs, this would assist in alleviating any inadequacies in the official response to misleading advertisements.
GRPI urges that the above recommendations be implemented so as to contribute to restraining unjustifiable representations and strengthening the response thereto.