Japan is about to write dark patterns into consumer law. Jiji Press reported on 10 September 2026 that the Consumer Affairs Agency (CAA) plans to regulate deceptive online designs under the Specified Commercial Transactions Act and expects to submit a bill to the ordinary Diet session next year. A separate expert panel has proposed banning the obstruction of subscription cancellations under the Consumer Contract Act. The CAA's goals are sound. What matters now is how the rules are drafted.
The case for acting
The strongest argument for regulation is that current law reaches only some tricks. According to Akita Sakigake Shimpo's report of 24 August 2026, the draft interim summary released that day says existing rules, mainly the Specified Commercial Transactions Act and the Act against Unjustifiable Premiums and Misleading Representations, restrict only certain techniques. Japan has no comprehensive dark pattern law. The draft defines the target as methods that use psychologically manipulative screen displays online to steer consumers into contracts. It would apply at three stages: advertising, contract formation and cancellation.
There is also evidence of harm. The same report cites an estimate from the Dark Pattern Countermeasures Association that domestic losses could reach about ¥3.2 trillion a year. That is an advocacy group's estimate and not an audited figure, so treat it as indicative. Even so, nobody disputes that a hidden recurring charge or an exit flow built as a maze costs real people real money. A reasonable regulator would want to act.
Japan has already tried part of this
This is not a blank slate. The CAA's own pages show the Specified Commercial Transactions Act was amended in 2021, and the agency now runs public campaigns on regular-purchase (teiki kounyuu) traps. It urges shoppers to save a screenshot of the final confirmation screen and to use the right to cancel a purchase offer made under a misrepresentation. It describes ongoing enforcement against deceptive regular-purchase schemes. These are targeted tools aimed at a specific, well-documented harm, and they point to a lesson. Rules tied to a concrete transaction step, such as the final confirmation screen, give businesses clear obligations and give regulators something to enforce.
The new proposal moves toward something broader. The Sakigake report says the draft contemplates a degree of comprehensive regulation because a list of individual tricks cannot keep pace with changing designs. Specific violations would then be set out in cabinet orders, ministerial ordinances and guidelines.
Where the risk lies
A general prohibition on psychologically manipulative design is hard to apply. Nearly all persuasive design uses some psychology. A prominent "best value" badge, a default selection, a countdown timer and a retention offer made before a cancellation all influence behaviour. Some are deceptive. Many are ordinary commerce that helps consumers compare options. If the line is drawn in guidelines after the fact, firms will not know where it sits. Small Japanese e-commerce sellers and startups, which lack in-house counsel, bear that uncertainty most heavily. Large platforms can absorb it.
Delegating the detail to ordinances and guidelines also brings a speech and innovation concern. Product interfaces are expressive choices, and an overbroad rule invites defensive design: bland, homogenized checkout flows that avoid any risk. Because the CAA must still define the line, the delegation is a drafting choice with real consequences.
What a proportionate version looks like
The cancellation proposal is the strongest part of the package. MLex reported on 1 September 2026 that the CAA expert panel proposed banning businesses from obstructing consumers who want to cancel subscriptions, amid a surge in complaints. Its draft interim report calls for amending the Consumer Contract Act, and its examples include withholding cancellation-relevant information and refusing cancellation requests. This is a clear, observable harm with a clear remedy. A consumer who signs up online should be able to leave online, and a regulator can check this by testing the flow.
Three design principles would keep the bill proportionate.
- Prefer outcome tests to design bans. Ask whether a consumer could clearly see the price, term and renewal conditions before committing, and whether cancellation took materially more effort than signing up. Do not ban categories of interface.
- Put the core definitions in statute. Where the Diet leaves key terms to ordinances, require public consultation and a published rationale before any new prohibited practice takes effect.
- Use evidence-based safe harbours. Offer model cancellation and confirmation flows, so that firms that follow them know they are compliant.
The CAA should also publish how it arrived at its figures. Loss estimates from advocacy groups are useful for agenda-setting but are a weak basis for drafting. Complaint data and enforcement records from the CAA and the National Consumer Affairs Center should anchor the impact assessment.
Why the sequencing matters
Two statutes are in play, and so are two panels. MLex describes the dark pattern work and the cancellation work as separate tracks, while the Sakigake report describes cancellation as part of the broader draft. That ambiguity needs resolving before the bill is drafted. If both laws regulate cancellation, firms face overlapping duties and inconsistent remedies. One clear rule for continuing contracts, with enforcement under a single regime, would serve consumers and businesses better.
Japan can still get this right. Of the proposals so far, the cancellation ban and clear rules for advertising and checkout screens are well supported by evidence. A broad catch-all prohibition, defined mainly through later guidance, is the part to scrutinize before it goes to the Diet in 2027.