Japan’s New Scam-Signal Network Tests Whether Defenders Can Move Across Borders

The fraud does not stay in one place long enough to belong to one company. A fabricated endorsement appears as a paid advertisement. A click leads to a social account, then to a closed messaging group. An apparent adviser builds confidence. A bank, payment service or cryptocurrency exchange becomes the final exit. Every handoff creates a gap between organizations—and every gap buys the scammer time.

Japan’s newest answer is to move fragments of intelligence across those boundaries before the criminal operation can move again. The Japan Anti-Scam Alliance, formally launched on August 26, is bringing technology companies, advertising and cybersecurity organizations, consumer specialists and international networks into a common framework. Its first test is narrow, but consequential: can one organization’s sighting become another organization’s early warning?

The decisive contest is measured in hours

A malicious URL may be obvious to an investigator who has already seen the campaign. It is just a new address to a second platform. A copied photograph, advertiser identity, domain, account or payment clue can acquire meaning when it is matched against information elsewhere. That is the premise behind a “signal”: not a verdict that a crime has occurred, but a machine- or analyst-usable indicator that can trigger scrutiny.

JASA says its model treats online fraud as a supply chain running from contact and redirection to settlement and remittance. Its own diagnosis is that defense organized service by service leaves openings that criminals exploit. The alliance describes itself as a neutral, nonprofit platform intended to connect telecommunications, advertising, finance, security, consumer groups, researchers and government. Those are institutional aims, not yet demonstrated outcomes.

Inside the Japanese pilot

The concrete initiative is smaller than the alliance’s broad ambition. JASA says that, beginning in August, scam-ad and scam-site signals supplied by the International Intellectual Property Protection Forum, or IIPPF, have been shared through the Global Signal Exchange, known as GSE. Participants are examining how the information can support cross-industry analysis and countermeasures.

JASA plans to test broader sharing involving fraudulent accounts, fake advertisements and scam websites. It also intends to convene councils and issue-focused working groups, study victimization and criminal ecosystems, develop best practices and work with public agencies. But the launch announcement does not identify a live blocking rule, a participating bank action, a completed takedown or an amount of loss prevented.

A coalition built around institutional differences

The Japan Trust & Safety Association, established in January 2025, operates the alliance’s secretariat. It listed four cooperating Japanese organizations at launch: the Japan Cybercrime Control Center; the Japan Interactive Advertising Association; the Nippon Association of Consumer Specialists; and IIPPF.

Google Japan and Trend Micro are founding support partners. Facebook Japan, OpenAI Japan, Match Group and TikTok Japan were listed as supporting partners. Overseas relationships were announced with the Global Anti-Scam Alliance, the Global Signal Exchange and the Global Coalition to Fight Financial Crime.

The list matters because those institutions observe different stages of a scheme. It must not be read as evidence that every organization contributes the same data, receives the same access or has adopted a common enforcement policy. JASA says supporting companies provide funding as well as knowledge and technical experience while respecting the platform’s claimed neutrality. No participant-by-participant obligations were published.

The losses explain the urgency—but not the pilot’s success

Japan’s National Police Agency recorded 5,893 cases of SNS-based investment fraud in the first half of 2026, with losses of ¥79.79 billion. Cases more than doubled from the same period a year earlier, while losses rose 126 percent. The agency calculated an average loss of ¥440 million per day.

Banner and similar advertising was the most common initial contact method for that category, the agency said. Losses linked to that route rose 252.5 percent year on year. For full-year 2025, the combined total for what were then classified as special fraud and SNS-based investment and romance fraud reached 43,000 recognized cases and ¥325.74 billion in losses.

There is a material source discrepancy worth recording. JASA’s Japanese announcement matches the police figure of ¥325.74 billion, while its English page says ¥324.1 billion. This report uses the primary National Police Agency figure. None of these statistics measures the pilot’s reach, and not every investment fraud began with an advertisement.

Japan is plugging into infrastructure launched in 2024

The Global Signal Exchange predates JASA. Google, the Global Anti-Scam Alliance and the DNS Research Federation announced it in October 2024 as an international clearinghouse for information about scams, fraud and other online abuse. GSE says its infrastructure lets qualifying organizations contribute and analyze indicators across sectors and services.

According to GSE, Google initially contributed more than 100,000 URLs associated with fraudulent merchants and ingested roughly one million signals. GSE reported that its holdings had expanded beyond 70 million signals by March 1, 2025. These are operator-published figures rather than an independent audit of unique, current or actionable records.

JASA has also agreed to begin dialogue with signatories to the 2026 Industry Accord Against Online Scams & Fraud, presented in connection with the Vienna summit. That voluntary accord emphasizes prevention, lawful information sharing, resilience and public education. It is not a treaty, a regulator’s order or a cross-border investigative agreement.

Government pressure is tightening the advertising gate

A separate Japanese initiative is pushing major platforms to harden the point where an advertiser enters the system. On August 7, the National Police Agency, Financial Services Agency, Consumer Affairs Agency, Digital Agency, Ministry of Internal Affairs and Communications, Ministry of Justice, and Ministry of Economy, Trade and Industry jointly sent requests to Google, LY Corporation, Meta Platforms, TikTok Japan and X.

The requests cover three areas. Platforms were asked to conduct reliable advertiser identity checks, including electronic methods; make such information as the advertiser’s name, location and verification status visible, along with disclosures about advertising and principally AI-generated material; and respond promptly to deletion requests involving ads suspected of violating law. The companies were asked to report implementation details by October 16 and specified operating results by March 16, 2027.

The documents are administrative guidance under Japan’s Administrative Procedure Act, not punitive dispositions. They are also distinct from JASA’s voluntary coalition. Government reporting demands, private signal exchange and each platform’s enforcement system may reinforce one another, but they do not form a single compulsory regime.

Cross-border sharing creates a governance problem of its own

The case for international exchange is straightforward. A fraudulent domain, image or persuasion script detected overseas may reappear in Japanese, and a Japanese sighting may warn services elsewhere. The legal and operational problem is deciding what can be shared, with whom, for what purpose and at what confidence level.

URLs change owners. Screenshots lose context. A legitimate advertiser, researcher, journalist or victim-support site can be misclassified. Adversaries can attempt to poison reporting systems. A low-quality signal copied quickly across services could multiply an error just as efficiently as a good signal multiplies protection.

The Industry Accord recognizes that anti-fraud work must be balanced with privacy and freedom of expression, and that information sharing must be lawful. JASA’s announcement emphasizes neutrality and the public interest. Yet the public material reviewed for this report does not set out a complete governance framework covering provenance, personal-data minimization, retention, access controls, correction, appeal, independent audit or remedies after an erroneous action.

The proof will be one interrupted journey

The easiest numbers for an alliance to publish are membership and signal volume. The more meaningful measures are harder: time from first report to detection elsewhere; the percentage of well-founded signals that lead to review or action; how often lawful content is affected; time to correct a mistake; and whether intervention occurs before a victim sends money.

This is also an economic question. Digital advertising depends on buyers believing that an apparent company or public figure is real. When fraud corrodes that assumption, legitimate advertisers lose attention, platforms and banks acquire heavier compliance costs, and consumers retreat from lawful transactions. Anti-scam infrastructure therefore protects market confidence as well as individual victims.

Japan.co.jp’s assessment is that the Japanese pilot deserves attention precisely because it is modest. It tests whether a clue collected in one institutional silo can become useful somewhere else. Its credibility will not come from describing collaboration as unprecedented or from counting logos. It will come when the alliance can document an interrupted fraud pathway without concealing its errors: an advertisement stopped, a payment delayed, a victim reached—and a mistaken flag corrected.