On April 1, 2026, Japan closed a gap that had sat in plain sight for two decades. Since the Act on Prevention of Unauthorized Use of Mobile Phones took effect in the mid-2000s, voice-capable SIMs sold in Japan have required identity verification — but data-only SIMs and eSIMs slipped through, because they carried no phone number to abuse. The Ministry of Internal Affairs and Communications (MIC) amended the Act's enforcement regulations (Ministerial Ordinance No. 167 of 2005) to require carriers to verify the identity of every data-SIM and eSIM contract holder using a passport, residence card, My Number Card or driver's licence (MIC press release). Carriers moved first: KDDI's povo switched its onboarding to the new verification methods on March 25, 2026 (povo official notice).
The case for the rule is genuinely strong
It is worth stating that case plainly before picking it apart. Japan is in the middle of a real fraud emergency. The National Police Agency reported in February 2026 that combined losses from special fraud, investment and romance scams reached a record ¥324.1 billion (about $2.1 billion) in 2025, up from ¥199.1 billion a year earlier, across roughly 42,900 reported cases. "Special fraud" alone — the impersonation scams Japan has fought for years — nearly doubled to ¥141.4 billion. Those 2025 figures followed a 2024 in which special-fraud losses had already surged 58.6% to ¥71.8 billion (Nippon.com).
Anonymous prepaid and data SIMs are a documented input to that machine. The loosely organised tokuryu crime groups behind much of the surge use disposable lines to recruit "dark part-time" (yami baito) labour, run SIM-swap attacks, and relay calls while staying unattributable. Aligning data SIMs with the verification regime that voice SIMs — and the carriers' record-keeping and the police's power to demand re-verification — have followed for years is internally coherent (MIC, framework overview). Tokyo is extending an existing rule, not inventing a surveillance apparatus from scratch. That instinct deserves credit.
But coherence is not effectiveness
Here the evidence is thinner than the rhetoric. The single largest component of Japan's 2025 losses was "fake police" scams — roughly ¥98.5 billion, about 70% of special-fraud losses. These succeed through social engineering: a convincing voice, a fabricated arrest warrant, a victim talked into wiring their savings. None of that depends on an anonymous SIM. The scam works on psychology, and the perpetrators increasingly sit offshore, beyond the reach of a Japanese onboarding form.
Determined criminals also adapt around attribution at signup — the one link a registration mandate touches. Stolen and rented identities, foreign-issued SIMs roaming on Japanese networks, and money mules all survive the new rule. International experience with mandatory SIM registration is sobering: across the dozens of countries that have imposed it, reviewers have found little robust evidence that it reduces crime, while it reliably raises the black-market value of stolen identities. Japan is tightening precisely the link in the chain that organised fraud finds easiest to route around.
The design problem: friction and centralisation
The rule's design compounds the concern. The amendment does not merely add data SIMs; it retires the lighter-touch verification methods carriers relied on — the "copy of an ID plus non-forwardable mail" route and the selfie-plus-photographed-document eKYC flow — and pushes verification toward reading the IC chip on an identity document, with the My Number Card and its JPKI authentication as the default online path.
For legitimate users, that adds friction at exactly the onboarding step Japan's MVNO and fintech sector spent years smoothing: short-term visitors, digital nomads, the elderly who never obtained a My Number Card, and foreign residents whose residence-card characters or addresses do not parse cleanly. For everyone, it deepens dependence on a single state credential for everyday connectivity — the kind of quiet centralisation that civil-liberties groups warn becomes hard to reverse once normalised (EFF, May 2026). A SIM database keyed to a national ID is both a standing target for breach and a standing temptation for scope expansion.
A proportionate version of the same rule
None of this argues for anonymous SIMs as a right. The targeted parts of Japan's reform are sensible and worth keeping: multi-factor matching before a second or third line, letting carriers refuse suspicious bulk-line orders, and preserving the police's existing power to demand re-verification of a flagged line. Those measures hit the actual fraud workflow — high-volume line acquisition — without burdening the ordinary customer.
The heavier-handed move is the blanket retirement of low-friction eKYC and the gravitational pull toward My Number. A proportionate version would keep at least one well-audited remote-verification path open for legitimate onboarding, guarantee passport-based verification for foreign visitors who will never hold a My Number Card, wall off the resulting identity data from secondary use with hard statutory limits and independent oversight, and commit MIC and the NPA to publishing whether the change actually moves the fraud numbers.
If special-fraud losses keep climbing through 2026 and 2027, that is evidence the fulcrum was wrong and the friction was paid for nothing. The danger is that Japan answers a social-engineering problem with an identity-infrastructure solution, accepts real costs to innovation and inclusion, and declares victory before the scammers even notice.