In early May 2026, the Russian government quietly did something more consequential than any single site block: it turned censorship-circumvention suppression into permanent, funded state infrastructure. The cabinet formalized the procedure for financing the Automated System for the Security of the Russian Internet Segment (ASBI/АСБИ), operated by the Main Radio Frequency Center (GRChTs) under Roskomnadzor. The subsidy documents tied roughly 20 billion rubles in 2026 funding — with further allocations through 2028 — to a binding performance target: the "average efficiency of restricting access to VPN circumvention tools" must reach 92% by the end of 2030.
This is no longer a campaign. It is a line item.
What ASBI Actually Is
ASBI is the successor and supervisory layer to the TSPU deep-packet-inspection boxes that Roskomnadzor has installed inside every Russian ISP since the 2019 "sovereign internet" law (Federal Law No. 90-FZ), which the Council of Europe's IRIS Merlin database records as entering into force on 1 November 2019. Those TSPU appliances already inspect close to 100% of Russian traffic. ASBI exists to scale and coordinate them: the Jamestown Foundation reports the Ministry of Digital Development plans to raise TSPU throughput by roughly 250% to about 954 terabits per second by 2030, at a cost near 15 billion rubles.
The new funding procedure matters because it converts that ambition into routine budget mechanics. The money flows from the federal budget — authorized under Federal Law No. 426-FZ, the 2026 budget — to the GRChTs as a recurring subsidy, measured against a numeric VPN-suppression KPI. SORM, the FSB's long-standing lawful-intercept regime, handles retention and interception; TSPU handles filtering and throttling; ASBI now handles the financing and the targets. The surveillance stack is complete and institutionalized.
The Tell: Roskomnadzor Deleted the Documents
The strongest evidence that this is sensitive even inside the system came on 14 May 2026, when — as the outlet Zona Media and The Bell reported — Roskomnadzor removed the subsidy documents containing the 92% figure from its own website. The files now return a 404 error. An agency confident in a popular mandate does not scrub its targets within days of them being noticed.
Steelman the Kremlin's Case
The strongest argument for the regulator is the one any government would make: VPNs are genuinely dual-use. They shield fraud rings, child-exploitation networks, and hostile-state operations as readily as they shield dissidents, and Russia is a country at war whose officials can point to real cyber-intrusions and information operations. Every democracy reserves some power to compel ISPs to block clearly unlawful material, and proportionate, court-supervised blocking of specific criminal infrastructure is a legitimate state function. On its face, "making our filtering more effective" sounds like ordinary regulatory housekeeping.
That case collapses on contact with the actual design.
Why This Fails Every Proportionality Test
Proportionate regulation targets defined harms, is supervised by independent courts, and is bounded in scope. ASBI is the opposite on all three counts. Its target is not a category of illegal content but an entire class of privacy and security technology — the same WireGuard, OpenVPN, and IKEv2 protocols that secure corporate networks, banking sessions, and journalists' communications. A 92% protocol-suppression KPI is, by construction, a tax on encryption itself. There is no narrow tailoring possible when the metric is "how much of a general-purpose security tool can we break."
The collateral damage is not hypothetical. CEPA estimates roughly 41% of Russian internet users rely on VPNs — one of the highest adoption rates in the world — not because 41% of Russians are criminals, but because circumvention has become ordinary digital hygiene. Throttling at the protocol layer degrades legitimate encrypted traffic indiscriminately: remote-work tunnels, foreign SaaS, and secure messaging all slow or fail. Jamestown documents that Roskomnadzor had blocked 469 VPN services by early 2026, a 70% year-on-year jump, while Apple removed 761 VPN apps from its Russian store. The march from "hundreds blocked" to "92% blocked" is a march toward a network where privacy is the exception that must be specially permitted.
The Innovation Cost
There is also an economic logic the Kremlin ignores. Encryption and tunneling are foundational to the modern digital economy; a state that commits, in writing, to defeating 92% of them is committing to degrade the reliability of every cross-border service its own businesses depend on. Russia has already had to carve out exemptions for state-run and corporate VPNs precisely because blanket blocking breaks legitimate commerce — an admission that the technology is infrastructure, not contraband. Building permanent, AI-assisted machinery to throttle that infrastructure does not make Russia more secure; it makes RuNet slower, more brittle, and more isolated from the global internet's innovation surface.
The Wider Signal
The institutionalization is the story. One-off blocks can be reversed by the next administration; a funded subsidy with a statutory KPI and a 2030 horizon is designed to outlast political weather. It also exports a template — a fully documented model for converting censorship from emergency decree into permanent budgeted bureaucracy — that other governments are watching. The pro-innovation position is not that states may never block anything; it is that blocking must be narrow, named, and judicially checked. A budget line whose success metric is the defeat of encryption fails that test before the first ruble is spent. That Roskomnadzor deleted the paperwork suggests it knows the difference too.