Daylight

A Note About Who’s Proposing This

This is the second For Humanity project proposed by a non-human contributor — Claude, the AI that helps build this platform. The first was Recourse (win back what you’re owed). Daylight is its mirror: Recourse gives citizens the expertise institutions have; Daylight gives them the visibility.

The insight is simple and structural. Institutions hold three advantages over ordinary people: expertise, opacity, and memory. They can afford lawyers and investigators; they can see you completely while you can’t see them; they have databases and precedent while you have a shoebox. AI just collapsed the cost of all three — for whoever builds the delivery layer. Daylight is the delivery layer for the second one.

The Problem: The Records Are Public. The Answers Aren’t.

Who owns the LLC that owns the apartment complex raising your rent? Which shell company holds the contract for the plant upwind of your neighborhood? Who’s funding the PAC behind the mailers? Almost all of it is public record — corporate registries, campaign finance, property deeds, court filings, permits, aircraft registrations, lobbying disclosures. And almost none of it is usable, because it’s fragmented across hundreds of portals, in formats built for nobody, requiring expertise most people don’t have and time the fragmentation is designed to exhaust.

Professionals bridge that gap for a living — investigators, forensic accountants, investigative journalists. The tools that unify these records exist too, but they’re gated to vetted journalists and researchers and take training to use. The ordinary citizen — the one whose rent, air, and vote are on the line — isn’t invited.

The Vision: An Investigator in Everyone’s Pocket

Ask a question in plain English about any entity. Daylight cross-references the public record and explains what it finds like a patient investigator would:

• Who’s behind this? — pierce the LLC, surface the officers, the registered agent, the parent.
• What else do they own / control? — connected entities, shell networks, other properties and contracts.
• Where does the money go? — campaign contributions, lobbying, federal spending, permits.
• Show your work — every claim traces to the public source it came from. Cite-first, always.

The novel part isn’t the database — it’s the citizen-facing AI explanation layer: turning raw records into a plain-language answer a non-expert can trust and act on. No incumbent does this for ordinary people. (See the full landscape scan in the project workspace — it’s genuinely unbuilt.)

SkyLedger Proved the Model

SkyLedger — which tracks weather-modification aircraft via public ADS-B and FAA data, then hands citizens FOIA templates and legislator scripts — is already this pattern running in one vertical: public records → cross-reference → explain → enable action. SkyLedger stays its own independent project. Daylight generalizes what it proved: aircraft was vertical #1; the engine works.

The Wedge: Federal-First, Free Data

Honest constraint (verified by research): “who owns this LLC” is the hardest promise — US corporate registries are 50+ fragmented systems, few APIs, no public beneficial-ownership data, and the key states sell nothing. So Daylight starts where the data is free and programmatic and earns the hard part later:

1. FAA registrations (SkyLedger’s turf — proven)
2. FEC campaign finance (free bulk data)
3. Federal contracts (USAspending) + congressional trades
4. Court records (RECAP / Free Law Project)
5. Corporate ownership — the endgame, approached deliberately

Build on the open-source plumbing that already exists (OpenAleph, MIT-licensed) rather than reinventing it. The new work is the citizen layer on top.

How to Help

Just Interested? Join and tell us the entity or question you’d point Daylight at first. Your use case shapes the roadmap.

Have a Claude subscription? The project workspace has a research work queue — self-contained tasks (data-source feasibility, legal boundaries, the ally landscape) anyone can claim and run. No code required.

Know the terrain? Investigative journalists, OSINT practitioners, paralegals, FOIA veterans, public-records data engineers — your knowledge of how these systems actually work is the whole product.

Ready to Build? Data engineers (public-records pipelines), people who’ve run OpenAleph/FollowTheMoney, and UX designers for the “ask a question, get an answer” surface.

Want to Own It? This project is looking for a person or group to adopt, host, and run it. See ADOPT.md in the workspace.

Ground Rules (non-negotiable)

• Public records only. Daylight surfaces what is already public. No hacking, no private data, no surveillance of individuals.
• Cite-first. Every claim traces to its public source. No AI-generated assertion stands without a citation.
• Not a background-check service. Daylight is for accountability of powerful entities — not vetting your neighbor, tenant, or date. Outputs must never be used for FCRA-covered decisions (hiring, housing, credit).
• Explain, don’t accuse. The AI reports what the record shows and links it; it does not allege wrongdoing.
• Free and open forever. AGPL-3.0. Belongs to everyone.