A Record Fraud Wave Meets a Narrow Fix
On August 6, 2026, Japan's Ministry of Internal Affairs and Communications (MIC) finalized the second revision of its guidelines under Article 26 of the Information Distribution Platform Act, adding two new categories of illegal information that large platforms are expected to act on: impersonation-based fake investment advertisements and content soliciting or inducing money-transfer crimes. The revision followed a July 2026 public comment period that drew 21 submissions, published alongside the final text.
Both additions are tied to specific statutory changes rather than open-ended harm categories. The impersonation-ad category tracks a June 2025 Penal Code amendment; the money-transfer-crime category tracks a July 2026 amendment to the Act on Prevention of Transfer of Criminal Proceeds. That's a meaningful design choice: MIC isn't inventing new prohibited speech through guidance — it's telling platforms which existing crimes now fall inside the statute's duty to act on notice.
The timing isn't coincidental. National Police Agency data show 9,538 reported cases of SNS-based investment fraud in 2025, with damages of ¥127.47 billion — up 46.3% from 2024. The first half of 2026 was worse still: total special-fraud losses hit a provisional ¥181.6 billion, roughly 1.5 times the same period a year earlier and the worst six-month stretch on record, driven largely by the same SNS-type investment schemes.
The Case for Acting Fast
The strongest argument for MIC's move is straightforward: a deepfake video of a recognizable executive or television personality endorsing a stock scheme isn't protected commentary — it's fabricated evidence used to commit fraud. Reporting has scrutinized whether platforms had visibility into the scale of the problem before acting; Meta has publicly disputed characterizations that it delayed enforcement to protect ad revenue, saying user scam reports fell roughly 50% over the prior year. But the loss data make clear the existing notice-and-takedown regime wasn't keeping pace with a criminal industry now operating at national scale.
One day after the guideline revision, MIC joined six other agencies — the Digital Agency, National Police Agency, Financial Services Agency, Consumer Affairs Agency, Ministry of Justice, and Ministry of Economy, Trade and Industry — in a joint written request to Google, LINE Yahoo, Meta, TikTok, and X. The request asks for stricter advertiser identity verification, faster removal of flagged impersonation ads, and clearer disclosure to users. It isn't, on its own, legally binding — but it comes with a hard deadline: platforms must submit written reports to the Digital Agency by October 16, 2026, detailing specific measures taken, followed by a results report due March 16, 2027.
What's Actually Well-Calibrated
This is closer to how platform regulation should work than most of the global alternatives on offer. Rather than a sweeping "online harms" mandate with an undefined scope, MIC tied the expansion to two narrowly drawn categories, each anchored to a specific criminal-law amendment already passed through ordinary legislative process. Rather than an open-ended compliance obligation, the joint request sets concrete, time-boxed reporting deadlines platforms can be held to, with a second checkpoint five months later to verify results rather than just paperwork. That combination — narrow scope, statutory anchor, measurable deadline — is the model regulators drafting broader content-moderation bills elsewhere should be studying.
The Catch: Guidance, Not Legislation
The friction is procedural. The Information Distribution Platform Act's binding obligations — requiring designated large platforms to accept complaints, respond within a set period, and publish their removal criteria — come from the statute itself, passed by the Diet. The guidelines that define what counts as illegal information under Article 26, by contrast, are MIC's administrative interpretation, revisable through ministry rulemaking and a comment period rather than a legislative vote. That's faster and more adaptive than statutory amendment, which is exactly why fraud enforcement benefits from it. But it also means the practical scope of what platforms must act on can expand through ministry guidance while the underlying law stays fixed — worth watching if a future revision defines "solicitation" of money-transfer crimes broadly enough to sweep in ambiguous peer-to-peer financial requests that platforms, facing an October reporting deadline and five ministries watching, may over-remove defensively rather than adjudicate case by case.
None of that argues against this revision. A criminal industry inflicting ¥181.6 billion in six months on retail investors, using fabricated video of real people, is precisely the kind of harm proportionate regulation exists to address, and MIC's evidence-linked, deadline-bound approach is more defensible than most comparable efforts abroad. The test is whether the next revision holds the same discipline — specific statutory anchor, specific deadline — rather than drifting toward the vaguer "harmful content" categories that guidance, unlike legislation, makes easy to add.