Temporal Authority Governance
for the Autonomous Era

Governing purpose

The infrastructure for a human-governed autonomous era

OCUP isn't another security product. It's an attempt to establish the Temporal Authority Governance infrastructure that increasingly autonomous societies can safely depend on — built through human-AI collaboration, independent verification, and governance architectures future generations can build on and improve.

Early partners aren't only evaluating today's evidence. They're helping shape the authority architecture of tomorrow.

The fuller case for what this looks like at civilization scale — including how it can extend into shared economic opportunity — is laid out at aicomsci.org.

What OCUP is

An authority layer, not an intelligence layer

OCUP — One Chip Unified Protocol — is a runtime authority evidence layer for autonomous systems. It doesn't replace an AI stack, a robotics controller, or a neural-control loop. It sits alongside them and answers a narrower question: was this action authorized, for how long, by whom, and what's the proof?

The mechanism is built on patent-pending temporal leases — bounded, expiring grants of authority — quorum-based approval for high-risk actions, controlled capability step-down as authority narrows, and an append-only audit layer.

Why it matters

Authority needs to be provable, not just intended

As autonomous systems take on higher-stakes actions — financial transactions, physical movement, infrastructure control — the question stops being "is the model good enough" and becomes "what was it actually authorized to do, and can that be proven after the fact."

That question matters to insurers pricing risk, to regulators drafting certification requirements, to fleet operators managing liability, and to any organization that needs an answer more durable than "we trust the model." OCUP's premise: authority should be a verifiable, time-bounded, externally governed property — not a policy the software agrees to follow.

Current evidence

What exists today, described plainly

The OCUP Pilot Portal hosts the current evidence set: an 18-scenario local reference test suite — 7 granted, 11 denied, 18 of 18 assertions passing — sourced from a specific commit in the reference control-plane implementation.

Executable Rust reference implementation. The current pre-production evidence harness and reference control plane are implemented in Rust to support deterministic scenario execution, memory-safe systems development, reproducible testing, and tamper-evident audit generation.

This is local, deterministic reference evidence: a controlled software harness demonstrating the authority logic, not a production distributed system. Rust is the implementation foundation for the present evidence layer; it is not, by itself, a production hardware-enforcement, certification, or safety claim. The current system does not reflect hardware-enforced deployment, live remote validators, or HSM-backed signing. Publication-readiness gating is local only at this stage.

Read the category thesis, benchmark model, industry applications, and paid-pilot pathway. No registration required.

IP & patent status

Two PCT applications, national security review complete

OCUP's core architecture — bounded temporal lease enforcement, quorum-based validator consensus, and the append-only audit layer — is covered by two Patent Cooperation Treaty (PCT) applications, filed to preserve an international priority date across jurisdictions.

Section 181 national security review has been completed, with a foreign filing license granted across three federal agencies, including the Department of Defense. No secrecy order.

A PCT filing establishes an international priority date — it is not a claim of granted patent status in any jurisdiction. Application numbers and filing dates available on request pending publication.

Pilot cohort

90-Day Self-Extension Denial Evidence Pilot

OCUP is opening a 90-day pilot cohort for organizations that want to test the core runtime-authority question directly:

Can an autonomous system preserve, renew, or extend its own authority after its approved runtime boundary expires?

The pilot is structured as an adversarial evidence engagement for robotics companies, insurers, fleet operators, regulators, and standards-adjacent reviewers. Participants define realistic operational scenarios, including lease expiry, attempted self-extension, network loss, stale or replayed authorization, and capability step-down conditions.

OCUP then produces a participant-specific evidence record showing what was authorized, what was denied, what expired, what stepped down, and what proof exists afterward.

Each participant receives a company-specific evidence report plus an anonymized cohort-level benchmark summary at the program’s close. The goal is not to claim a finished production-certified product. The goal is to create a documented technical and commercial record around OCUP’s boldest deliverable:

Proof that an autonomous system cannot preserve, renew, or extend its own authority beyond its approved runtime boundary.

QVN / Partnerships horizon

From local authority evidence to network-verified human governance

OCUP's current evidence layer is local and deterministic by design. The next horizon is QVN — the QSAFP Validator Network — a distributed quorum layer for lease validation, authority continuity, and human-governed reauthorization at fleet scale.

OCUP is seeking technical partners in robotics control, HSM-backed signing, edge attestation, distributed validators, insurance telemetry, and standards-aligned safety evidence.

The mission is simple: local enforcement today, network-verified human authority tomorrow — preserving humanity's seat at the table while enabling safe, scalable autonomous systems.

That validator network, as it matures, is designed to become more than an enforcement layer — a channel through which people, not only corporations, can participate in and hold a durable stake in governing the systems that increasingly act on their behalf. The fuller economic case is laid out at aicomsci.org.

This describes design intent for the validator network's evolution, not a current compensation program or income guarantee.

Stay informed

Get critical updates

Material changes to the evidence boundary — new benchmark results, validation status changes, production-path milestones. Sent only when something actually changes.

Email address Subscribe