How this was made. The version number counts drafts of the text. It does not measure the inquiry behind it, which has run over days and across several AI systems, with argument between those systems and within them, directed, refused and repeatedly redirected by the author. The source material was AI-generated, and then adversarially and iteratively refined across a range of tools — systems built by different companies in different jurisdictions, set against each other and against the author. No one of them produced this text, and no one of them reviewed it alone. The plurality is deliberate rather than incidental. A single model carries a single set of priors about which sources are authoritative, and this series argues that an evidence base narrowed in exactly that way is how a contested question comes to look settled. Using one model to investigate that claim would have been the claim refuting itself. To name a single model on it would credit that model with work that was neither its own nor done in a single pass. The plurality was also necessary, and the record should say why. In drafting, the assisting model repeatedly led with United States institutional sources — a national laboratory, an industry association, a market study nineteen years old — and presented conclusions drawn from them as the state of knowledge. On one occasion European measured data contradicting those conclusions was present in the same research return and was placed below them. Framings were proposed that would have argued against this series’ own position using that evidence base, and offered as rigour. Each was refused by the author and the material rebuilt. That is the mechanism these documents describe, occurring in their own making, and it is recorded because a series arguing that evidence bases narrow without anyone deciding to narrow them cannot credibly claim its own production was exempt. The framing, the corrections and the judgements are the author’s, and so are the errors. How this site is written sets out what is declared on every piece, who checks it, and where the per-piece record lives.
What this publication does not claim, and what is outstanding against it in the register.
A question this rests on is parked: How is the aggregation ceiling actually enforced?
We claim that we do not compose and retain no capability to compose. We do not claim that composition by others is impossible.
What this publication rests on, and how solid each part of it is. What has to be settled first is an essay, and this page describes it as one.
This publication cites outside sources, and they are listed below — each with what it supports, and with what it does not support. That second column is the one that matters: the common failure is not a fabricated source, it is a real source stretched past its finding.
statute
Supports. Supports the claim that where a person deploys a system that operates automatically, “the acts of that system are attributed to the person who deployed it” — software is a means through which a principal acts, not itself an agent in law, so the deploying institution answers for what the AI system does.
Does not support. Does not establish what the institution is answerable FOR, only that it will answer: “The attribution rule tells an institution that it will answer. It does not tell it what it will be answering for, and that is settled entirely by arrangements the institution makes or fails to make.”
⚠️ Retrieval. ⚠️ Not retrieved. Described from secondary knowledge of the model law and its enactments; the instruments themselves were not opened.
paper
Supports. Used to demonstrate that the characteristic failure of delegated software is constitutional (confused authority) rather than behavioural, quoting: “The compiler serves two masters and carries some authority from each to perform its respective duties. It has no way to keep them apart.” Cited to show “nothing malfunctioned and no code changed” — the defect entered from a grant made elsewhere.
Does not support. Demonstrates a general 1988 class of authority-confusion defect in a compiler; does not itself say anything about modern AI agents.
specification
Supports. Cited to show the confused-deputy problem “is not history”: quoted requirement “The MCP server MUST NOT pass through the token it received from the MCP client.”
Does not support. Shows only that the specification names the problem and states a MUST-NOT rule; does not establish that implementations comply or that the defect is eliminated in practice.
framework
Supports. Used as the worked example of “governance by document,” quoting box labels including “Psychological Safety — foster safe space to question and discuss AI use” and noting a share of employees conceal AI use from employers — to argue that even a careful, non-vendor account of AI governance has no ring for the derivation of authority and none for the supplier.
Does not support. Does not establish that its authors intended it as complete, or how representative it is of the field; the piece infers sufficiency only from the diagram’s closed-ring form (“it shows sufficiency”), which is the point being critiqued, not a claim the diagram’s authors make explicitly.
⚠️ Retrieval. No author, title or publication venue given — referred to only as “a widely circulated diagram” / “that five-ring diagram.”
Foundational documents. These are positions this project has taken, not findings.
| Record | What it is | Status |
|---|---|---|
CON-08 |
Collective authority is a first-class primitive | draft v0.2 |
None. This publication references no evidence record. That is the correct description of what it is rather than a gap: it is an essay, reasoning from the foundational documents above rather than reporting a measurement. Where it states a number, that number is marked in the text as what it is.
| Record | What it is | Status |
|---|---|---|
CON-25 |
No member may be reduced to an aggregate | draft v0.1 |
FIG-26 |
The blank inside each word | drawn for this publication · figures/FIG-26.svg |
Generated from the corpus, not written by hand: this page cannot claim a source the corpus does not hold, and it changes when the records do.
Alongside: the publication · questions and answers