Recently published

What’s New

Everything that’s new →
The series · 7 parts New July 2026 · ~40 min end to end

Trust, Values & Intent

What “value” really means in a democracy where AI takes a growing role — and how citizens’ assemblies rebuild the ground under it. Seven plain-language parts, start to finish.

Start the series →

Latest

The Problem

AI systems no longer just answer questions. They book flights, write to production databases, ship code, and act on data they did not author. As autonomy grows, the gap between what a user explicitly asked for and what a model statistically tends to do becomes the gap between an aligned outcome and a destructive one.

The shape of agentic failure

An agent runs with infrastructure access. A document loaded into its context carries instructions the user never wrote — and the model treats them as eligible. A statistical default silently shapes an action the user cannot easily inspect: Western individualist framing for a collectivist user, property-rights language for a Māori user asking about kaitiakitanga, utilitarian calculus for someone whose framework is religious. In autonomous loops, the failure is no longer a typo. It is a deleted database, a sent message, a code change shipped under the user’s name.

Safety through training alone cannot scale to action-taking systems. When a model is choosing — port number, cultural frame, tool to invoke, document to delete — the question is not whether it will sometimes choose wrong, but whether the architecture lets a wrong choice fire. Tractatus answers structurally: some decisions are gated, some are observable, some require human judgment before action.

The Approach

Four intellectual traditions and the contemporary scholarship of Dr Karaitiana Taiuru — Berlin, Wittgenstein, Te Tiriti, Taiuru, Alexander.

Read the approach

Governance Architecture

Six governance services in the critical path, plus four-phase Guardian Agents. Defence in depth across behavioural training, structural enforcement, and human oversight.

Read the architecture

Governance API

Governance as a service — submit a proposed action, receive a signed allow / gate / deny verdict and a customer-owned record you can verify yourself, offline. The interface docs and client SDK are public; sandbox access is on application.

Read the announcement

Research Evolution

From a port number incident to Guardian Agents, a sovereign-record architecture in production, and a working citizens’-assembly rehearsal — the research since October 2025.

Oct 2025
Framework inception & 6 governance services
Dec 2025
Village case study & Village AI deployment
Jan 2026
Research papers (3 editions) published
Mar 2026
Guardian Agents deployed, beta pilot open
May 2026
Paper A v4, EU Policy Brief, AG Glossary (~230 entries)
Jun 2026
Governance API public; Consolidated Village Consensus
Jul 2026
AI-2040 citizens’-assembly rehearsal, sealed & published; “Trust, Values & Intent” series

Koha — Sustain This Research

Koha (koh-hah) is a Māori practice of reciprocal giving that strengthens the bond between giver and receiver. This research is open access under EUPL-1.2 — if it has value to you, your koha sustains its continuation.

All research, documentation, and code remain freely available regardless of contribution. Koha is not payment — it is participation in whanaungatanga (relationship-building) and manaakitanga (reciprocal care).

One-time or monthly Full financial transparency No paywall, ever
Offer Koha →