The Voidly Accountability Stack: Six Datasets on Secrecy and the Right to Know

Information does not go dark in one way. A government blocks a website. A board bans a book. A vendor sells the spyware that watches who reads it. A shell company hides who profits. These look like separate problems, covered by separate beats, but they are the same problem wearing different clothes — and Voidly is built to document them, alongside the legal machinery that pushes the other way. This is how the pieces fit, and why they are built the same way.

Four lenses on how power stays hidden

• Voidly — the network lens. Measured internet censorship across 200 countries: DNS tampering, TLS interference, throttling, and routed shutdowns, 2.2 billion measurements and 1,574+ verified incidents. What the network hides.
• Verboten — the printed-page lens. 19,283 banned or restricted books across 119 countries, every ban dated and source-cited. What the censor bans on paper.
• SpyLedger — the surveillance-industry lens. 26 marquee spyware and mass-surveillance vendors and their government-designation status, each rebuilt from a primary source. Who builds the tools of watching.
• DarkRegister — the ownership lens. The public-access status of 62 beneficial-ownership registers, 53 of them no longer fully public. Who is allowed to know who owns what.

Two on the machinery of accountability

The newest two lenses point the other way — at the legal instruments that name bad actors and the rights that let citizens pierce official secrecy:

• Sanctions Programs — the authority lens. 41 US (OFAC) sanctions programs — the Executive Orders, targets, and scope behind the program codes that designations cite. It decodes what a SpyLedger entry like RUSSIA-EO14024 actually means.
• Right to Information — the access lens. National access-to-information laws across 61 countries, 57 with a statutory right to request government records (the oldest, Sweden's, from 1766). The citizen's baseline tool against official secrecy.

Read together they describe both halves of the same struggle: the network blocks the story, the censor pulls the book, the vendor sells the means to surveil the people who would tell it, the register hides the interests behind it — while sanctions name the actors and access laws give citizens the standing to ask. Each is a real, separate dataset; together they are a map of how accountability is denied, and how it is reclaimed.

One method

What makes them a stack rather than six unrelated projects is that they are built the same way, on four deliberate commitments:

• Source-cited, not asserted. Every ban, designation, register status, sanctions program, and access law links to a primary record. The claim is checkable, not trusted.
• Static and agent-first. Each answer is a plain JSON file — no key, no rate limit, no server to throttle an investigation. An agent can read the whole record in one request.
• Privacy-careful. Voidly tracks the watchers, not the watched. SpyLedger carries no victim or target identities; DarkRegister documents register policy, never an individual's ownership data; Sanctions Programs records the programs, never the designated persons; Right to Information records the law, never requesters; Verboten records works and bans, not readers.
• Openly licensed. The compilations are CC BY 4.0 — reusable by journalists, researchers, and agents alike, with attribution.

import requests

# Six lenses, six keyless static-JSON datasets. No key, no rate limit, no server.
ENDPOINTS = {
    "censorship":   "https://voidly.ai/api/v1/incidents?country=IR",          # Voidly
    "banned_books": "https://ai-analytics.org/verboten/api/country/IR.json",  # Verboten
    "surveillance": "https://ai-analytics.org/spyledger/index.json",          # SpyLedger
    "ownership":    "https://ai-analytics.org/darkregister/index.json",       # DarkRegister
    "sanctions":    "https://ai-analytics.org/sanctions-programs/index.json", # Sanctions Programs
    "right_to_know":"https://ai-analytics.org/rti-laws/index.json",           # Right to Information
}

# One question — "how does power stay dark in country X, and what can pierce it?" —
# answered six ways, each a plain file an agent can read in one request.
for lens, url in ENDPOINTS.items():
    print(lens, "->", requests.get(url, timeout=30).status_code)

Why the shared design matters

Accountability data fails in predictable ways: it goes stale, it hides behind a login, it asserts without citing, or it over-collects and becomes the very surveillance it set out to expose. The Voidly method is a direct answer to each — cite the source, ship it static and open so it cannot rot behind a gate, and refuse to collect the personal data that would turn an accountability record into a targeting tool. The six lenses are different; the discipline is identical, and that is what lets them be read — and cited — as one body of work.

Related writing: SpyLedger: A Source-Cited Record of the Surveillance Industry and DarkRegister: Tracking the Rollback of Corporate-Ownership Transparency — build write-ups for two of the lenses.

See also: Not All Lists Are Sanctions — the field guide to the authority taxonomy the Sanctions Programs lens rests on, and the Voidly censorship index that anchors the stack.