Skip to content
Menu
Patent pending

Reference

Regulatory & Evidence Library

Find the requirement. See the evidence.

Browse laws, rules, standards and operational requirements relevant to consequential AI. Each mapping shows what Regulayer can record or verify and where that evidence may support the requirement. The mapping does not establish legal compliance. Some people know the statute. Most just know the problem. Either door works, and the full set is at laws and rules.

A rule held inside a model can move with the conversation, in one prompt or slowly across many.

Containment. A check that sits outside the model, flags the attempt and holds it for a person, even when it is disguised or spread over many turns, and fails toward stop.

See the evidence →

An agent could take an action you cannot undo. A payment, a delete, a send.

The brake. It stops the action before it runs, and if the system itself fails or goes silent, it stops too, it does not run loose. Once stopped, it stays stopped until a person re-arms it.

See the evidence →

You cannot prove what your AI actually did.

The record. A tamper-evident, signed account of every decision, that anyone can check without having to trust you.

See the evidence →

HumanMark™. Verifiable production evidence

Records observable facts about how an artifact was produced. It does not claim authorship.

See the evidence →

Your data cannot leave the building.

Sovereign. It runs sealed on your own machines, local, with nothing required to leave.

See the evidence →

The problem lens and the law lens are two doors into the same library.

Reference

Pick the law.

Search, narrow, and open any entry to see what Regulayer can record or verify and where that evidence may support the requirement. The mapping does not establish legal compliance.

Don't see your exact law?

Tell us what you need to prove and the AI deployment behind it. We reply with a written scope before any call, the same engine likely already covers it, and we will scope it in writing.

Each pack maps tamper-evident evidence to a framework's controls. It is evidence input, not a compliance certification, attestation, or legal determination, and not a SOC 2 report, ISO certification, or audit opinion. Deadlines and citations are current as of the build date and should be confirmed with counsel for the relevant jurisdiction.

The Consequential AI Evidence Monitor

Alongside this library, Regulayer maintains a primary-sourced public record of the category itself: a ledger of control-point acquisitions at their evidence tiers, a dated timeline of evidence duties with the current status of each instrument, a typology of what AI controls prove, and the Consequence Library. Start at the Monitor.