Nagoya Man Accused of Using 124 Accounts to Claim 3.7 Million Yen in Food Delivery Refunds

Police trace 1,095 orders to a suspected refund scheme

A 16,000 yen refund for ice cream, bentos and chicken steaks forms part of a much larger fraud investigation in Nagoya, Aichi Prefecture. Police say 38-year-old Takuya Higashimoto repeatedly received food ordered through Demae-can, ate the meals and then claimed that the deliveries had never arrived.

Investigators suspect the scheme involved 1,095 orders over roughly two years, with losses exceeding 3.7 million yen, about US$24,000. Higashimoto was arrested in early October on suspicion of fraud. Police say he used 124 accounts, false names and addresses, and prepaid mobile phone cards to make repeated claims while avoiding links between his profiles.

The case centers on contactless delivery, which allows a courier to leave food at a customer's door without meeting them. That convenience, widely used since the pandemic, also leaves platforms with a difficult task when someone disputes a delivery: deciding whether the food went missing or the complaint is false.

Demae-can has said it will strengthen identity verification and introduce alerts for abnormal trading activity. Those changes address two parts of the alleged scheme: the repeated creation of accounts and the pattern of refund requests spread across them.

As of October 16, 2025, the case remained under investigation, with prosecutors to determine the final charges. The arrest has drawn intense public reaction in Japan and neighboring markets using similar services. Some users have focused on refund rules that appear easy to exploit; others have expressed frustration at the persistence involved in managing so many accounts for so long.

How a delivery became a refund request

Police describe a straightforward process repeated through the Demae-can app. Higashimoto allegedly ordered food, selected contactless delivery and then used the app's chat function to report that the order had not arrived. Investigators say he had actually received the meals.

A transaction highlighted by authorities took place on July 30. Higashimoto allegedly opened a new account using a fictitious name and address, ordered ice cream, bentos and chicken steaks, and received the food. He then claimed it had not been delivered, prompting a refund of 16,000 yen, about US$105, that same day.

Investigators say he often selected more expensive items, including eel bento, hamburger steaks and premium desserts such as ice cream. Higher order values meant larger refunds from the same type of false complaint.

Across the suspected scheme, the losses of more than 3.7 million yen spread over 1,095 orders amount to an average exceeding 3,379 yen per order. The July 30 refund of 16,000 yen illustrates how a single claim could involve considerably more money than that average.

Why contactless delivery complicates disputes

Contactless delivery serves busy couriers and customers who prefer privacy. A driver can leave an order at the door and move on without waiting for a personal handover. The same arrangement reduces the opportunity for both sides to confirm receipt together.

Many delivery services ask couriers to photograph food at the delivery location. Customer service teams may still grant refunds when food disappears from a porch or a customer insists that an order never appeared, even if a photograph exists.

Platforms therefore have to weigh delivery evidence against a customer's complaint. A genuine customer may need a quick refund for missing food, while someone making a false claim can try to take advantage of the same process. Generous refund rules are intended to preserve customer trust, but repeated abuse can turn that convenience into a financial loss.

How 124 accounts fragmented the record

Authorities say Higashimoto managed 124 Demae-can accounts from April 2023, often registering with false names and incorrect addresses before cancelling them within days. He also allegedly bought prepaid mobile phone cards using fake credentials.

Fresh phone numbers and account details made refund requests appear to come from different customers. By spreading orders across many profiles, he limited the number of suspicious claims attached to any one account and reduced the likelihood of an automated block.

Short account lifespans also leave platforms with less information about a customer's behavior. A profile may disappear after only a few orders and refund requests, before enough activity accumulates to trigger closer scrutiny.

Retail and delivery platforms commonly assess several signals together: previous refunds, the device or browser used, delivery addresses, payment methods and feedback from drivers. Some use rules and machine learning, software that looks for patterns in data, to identify claims that appear risky.

Device fingerprinting is another tool. It uses characteristics of a phone or browser to help connect activity across accounts. Location patterns can also help identify related transactions. Repeated changes to names, phone numbers and account details make those connections harder to assemble.

When records remain divided among many profiles, a system can see numerous apparently separate customers instead of one repeat user. Connecting those records may then require a manual investigation, adding time and expense.

The dates behind the investigation

The suspected activity stretched across roughly two years before the early October arrest. Police have identified the start of the account activity and a specific July transaction within that period.

• April 2023: The period during which investigators say Higashimoto managed 124 accounts began.
• July 30, 2025: He allegedly obtained a refund of 16,000 yen after receiving an order containing ice cream, bentos and chicken steaks.
• Early October 2025: Police arrested Higashimoto in Nagoya on suspicion of fraud.
• October 16, 2025: The investigation remained ongoing, with prosecutors to determine the final charges.

Police say Higashimoto had been unemployed for several years and lived in Nagoya. During questioning, according to police, he described beginning with a test before continuing as the refunds accumulated.

Higashimoto, the suspect in the investigation, gave this explanation of his behavior, according to police:

At first, I just tried this trick. I couldn't stop after reaping the rewards of my fraud.

His reported explanation follows a pattern officers encounter in fraud that requires little technical complexity: an initial success encourages repetition, followed by escalation as the person grows confident that the system will not flag the activity. Such schemes can continue until a courier challenges a complaint, an analyst connects transactions or law enforcement intervenes.

What Demae-can says it will change

Demae-can has said it will strengthen identity checks and develop alerts for abnormal trading activity. The measures are aimed at identifying repeat abuse even when it is distributed across multiple accounts.

Stronger identity verification can involve checking names, addresses and phone numbers against public records, or requiring authentication through two steps during registration. These are examples of how tighter checks can work; the company's stated response centers on stronger verification and alerts.

Alerts can identify a high frequency of refunds, unusual ordering hours or rapid account creation connected to the same device or network. Combining those signals gives a platform more information than simply counting complaints on each individual account.

Delivery services also test changes to the handover itself. These include requiring more photographic evidence, asking customers to provide a delivery code for orders assessed as riskier, or temporarily limiting refunds for accounts with repeated complaints.

Each measure creates a practical tradeoff. More checks can make ordering and resolving problems slower for legitimate customers. Fewer checks leave more room for abuse. The challenge is to direct additional scrutiny toward suspicious activity while keeping ordinary deliveries and genuine complaints manageable.

Refund abuse extends beyond Nagoya

A fraud prevention firm working with major delivery platforms estimates that refund claims account for roughly 48% of consumer fraud on delivery apps. That figure places refund abuse close to half of the consumer fraud covered by the firm's estimate.

Online forums circulate instructions on wording complaints, obtaining fresh email addresses and phone numbers, and returning to platforms after a ban. Some schemes also repeatedly claim discounts intended for new users by opening multiple accounts.

Those discount schemes consume marketing budgets and inflate growth statistics. A business may appear to be attracting new customers when the activity actually comes from existing users repeatedly presenting themselves as newcomers.

The wider retail sector faces related losses. A joint report by Appriss Retail and Deloitte estimated US$103 billion in fraudulent returns in 2024. The figure covers retail returns, while the Nagoya investigation concerns food delivery refunds.

In Jiangsu, China, police confronted a group of three people who survived for weeks with only 19 yuan by rotating claims of missing deliveries across accounts. The amount was much smaller than the suspected Nagoya losses, but the repeated use of separate accounts followed a similar pattern.

Some delivery services respond by requiring customers to share a code with couriers to confirm receipt. Others apply closer scrutiny when several complaints arrive in a short period or a device raises concerns about its location or previous activity. Changing fraud tactics keep pressure on platforms to update their detection models, policies and staff training.

Who bears the cost of false complaints?

Restaurants can lose money through ingredients, preparation time and delivery commissions when orders are refunded. Businesses operating with narrow profit margins may find that a series of disputed orders erases a day's profits.

Some restaurant partners report more disputes during busy periods, when employees have less time to preserve evidence and challenge refund decisions. A meal still has to be purchased, cooked and packed, even when the payment is later disputed.

Couriers face a different risk. Repeated complaints that food never arrived can lead to restrictions or deactivation of a driver's account, including cases in which the driver followed the delivery instructions.

That pressure can encourage couriers to insist on meeting customers even when contactless delivery was requested. The result is less convenience for customers and a change to the service they selected.

Honest customers can also encounter stricter refund rules, more demands for delivery photographs or requirements to provide confirmation codes. People whose food genuinely went missing may face longer waits and more screening before receiving their money back.

For restaurants, drivers and customers, the central issue is how a platform distinguishes a real delivery failure from a false complaint. The decision affects who absorbs a loss and how much evidence everyone must collect during an otherwise routine order.

The legal case and the evidence trail

Higashimoto was arrested on suspicion of fraud. Under Japan's Penal Code, fraud can carry a prison term of up to 10 years. Prosecutors will decide the final charges as the investigation proceeds.

Investigators in similar cases sometimes also consider obstruction of business by fraudulent means, an offense used when deceptive conduct disrupts business and causes financial harm. The legal assessment depends on the conduct established through the investigation.

Connecting many transactions to one person requires a detailed record. Delivery platforms can provide order histories, chat logs, refund decisions, device identifiers and courier notes. Together, those records can show how orders were placed, what customers told support staff and how refunds were approved.

Phone carriers can help confirm ownership of numbers used to register accounts. That information can support efforts to connect profiles that appeared separate within an app.

A careful evidence trail supports prosecution while reducing the risk that genuine customer disputes become entangled in the investigation. In a case involving 1,095 orders and 124 accounts, linking activity across records is a central part of establishing what happened.

What a stronger refund system involves

Platforms have several ways to reduce repeat abuse without making every customer complete the same lengthy checks. A layered approach combines identity information, device information and patterns of behavior, with closer review for accounts or orders assessed as riskier.

The practices becoming common across the industry address different stages of an order, from account creation to a disputed delivery:

• Verify phone numbers and addresses at registration using trusted data sources.
• Use device fingerprinting to connect suspicious activity across accounts using the same phone or browser.
• Require delivery confirmation codes for riskier orders and clearer photographs for contactless deliveries.
• Limit how many accounts can be created from the same device or network within a short period.
• Flag unusual refund patterns automatically and refer suspicious cases more quickly to staff for review.
• Work with restaurants and couriers to preserve delivery evidence, including clearer notes and photographs showing when they were taken.
• Apply bans across the network that continue when serial abusers change devices or identity details.
• Explain refund policies to customers, including the evidence needed when a delivery genuinely goes wrong.

These measures work on different weaknesses. Account limits address rapid registration, delivery codes add confirmation at the handover, and connected records help identify repeated activity hidden behind fresh profiles.

Platforms that combine identity, device and behavior data tend to identify repeat abuse faster. The purpose is to preserve refunds for customers with real problems while directing serial abusers toward more rigorous checks. The Nagoya investigation shows how much can accumulate when repeated claims remain divided among accounts for an extended period.

Key Points

• Police arrested Takuya Higashimoto, 38, in Nagoya in early October 2025 on suspicion of fraud involving Demae-can refunds.
• Investigators suspect 1,095 orders over roughly two years caused losses exceeding 3.7 million yen, about US$24,000.
• Police say he used 124 accounts, false names and addresses, and prepaid mobile phone cards to avoid detection.
• A July 30 order allegedly led to a refund of 16,000 yen after the food had been received.
• Demae-can says it will strengthen identity verification and introduce alerts for abnormal trading activity.
• False refund claims can create losses for restaurants, account penalties for couriers and tighter checks for legitimate customers.
• Fraud under Japan's Penal Code can carry up to 10 years in prison. Prosecutors were still to determine the final charges as of October 16, 2025.