Between June 1 and June 3, 2026, three different Istanbul criminal courts of peace issued three separate access-blocking orders in response to a single investigative report. The first, from the 9th Criminal Court of Peace, blocked Kısa Dalga's five-part "Visa Empire" (Vize İmparatorluğu) series by journalist Canan Coşkun, which examined how VFS Global's Turkish subcontractor, Gateway Management, allegedly dominated the country's visa-appointment market and exposed applicants' personal data through its "Premium Lounge" service. The second, from the 5th Criminal Court of Peace, blocked news coverage and social-media posts about that first block. The third, from the 7th Criminal Court of Peace, blocked coverage of the first two blocks. Each order cited the same legal basis: Article 8/A of Law No. 5651, Turkey's internet law, invoking "national security" and "public order." On June 15, 2026, the International Press Institute condemned the pattern, calling on Ankara to "end the use of arbitrary court orders citing bogus 'national security' and 'public order' concerns to censor journalistic social media accounts and news content" (IPI).
What the investigation actually found
The "Visa Empire" series was Turkey's contribution to a larger cross-border investigation coordinated by Lighthouse Reports across 14 outlets in 12 countries. Coşkun's reporting alleged that applicants paid intermediaries up to €300 to secure Schengen visa appointments, that rejection rates climbed from 3.9% in 2015 to 14.6% in 2025, and that Gateway's owner, Halis Ali Çakmak, had business ties to figures connected to former Foreign Minister Mevlüt Çavuşoğlu (Stockholm Center for Freedom; bianet). None of that is a national-security matter in any conventional sense. It is a consumer-protection and corruption story about a private contractor's fee structure — the kind of reporting that, in most democracies, triggers a regulatory inquiry, not a court order erasing it from the internet.
The steelman: emergency blocking isn't inherently illegitimate
Before dismissing Article 8/A outright, it's worth stating its strongest justification fairly. Every serious internet-governance regime, including the EU's own Terrorist Content Online Regulation, allows platforms and authorities to remove certain content within one hour when the harm is genuinely irreversible: live-streamed violence, child sexual abuse material, doxxing that endangers a named individual. A court process that takes weeks cannot stop harm that unfolds in minutes. Article 8/A's structure — an emergency block followed by mandatory judicial confirmation within 24 hours and automatic expiry if a judge doesn't rule within 48 — mirrors that logic on paper. Fast first, review second, is a defensible sequencing when the underlying category of harm is narrow and genuinely urgent.
Where the Kısa Dalga case breaks that logic
The problem is that Turkish courts are no longer confining Article 8/A to that narrow category. A dominant visa broker's fee schedule is not a national-security emergency. And the second and third orders make the abuse undeniable: blocking coverage of a blocking order cannot, under any reading, protect national security or public order, because the existence of a court order is a public fact about the state's own conduct, not a secret whose disclosure endangers anyone. What the cascade actually protects is the reputational comfort of whoever objected to the first article. Reporters Without Borders' Turkey representative, Erol Önderoğlu, called the sequence "the tragicomic level censorship has reached in Turkey" — a fair description of a legal doctrine turned self-referential.
This is not an isolated incident. IPI's June 15 statement ties the Kısa Dalga cascade to the April 2026 blocking of Cumhuriyet's X account, which cut off roughly 3.5 million followers, plus orders against Mezopotamya Ajansı and JINNEWS. And the numbers show a tool scaling far beyond emergency use: according to the Freedom of Expression Association's (İFÖD) June 2026 report Digital Martial Law, Article 8/A decisions rose from 71 in 2024 to 179 in 2025, with the internet addresses they targeted jumping from 828 to 6,300 — a nearly eightfold increase in one year (Stockholm Center for Freedom). Total website and domain blocks in Turkey reached roughly 232,000 in 2025 alone, pushing the cumulative count since 2018 past 1.5 million. A provision built for hour-scale emergencies is now running at the volume of routine administrative censorship.
The judicial "safeguard" isn't doing its job
Article 8/A's design assumes the 24/48-hour judicial check will filter out overreach. In practice, three different magistrates approved three consecutive, increasingly absurd orders in three days without any of them apparently asking whether a corruption story about a visa contractor met the statute's own "life, safety, national security" threshold — let alone whether banning coverage of a ban could ever meet it. IPI's specific demand — that Turkey "publish all blocking decisions in full to enable legal defense" — targets the actual failure point: opaque, unreasoned orders make judicial review a formality rather than a check, because neither the public nor the outlet involved can contest reasoning that was never disclosed.
What proportionate reform looks like
None of this requires Turkey to abandon expedited blocking as a tool. It requires narrowing Article 8/A back to the catalog of genuinely time-critical harms it was written for, publishing the reasoning behind every order so magistrates and the public can evaluate whether "national security" is being invoked in good faith, and adding an explicit carve-out barring the use of the statute against reporting on matters of public interest — including reporting on the statute's own use. A law that can be turned against coverage of itself has stopped functioning as an emergency measure and become a mechanism for insulating power from scrutiny. Fixing that scope problem, not eliminating rapid-response blocking altogether, is the proportionate answer.