What's new · Government and defense collateral

Government and Defense Decisions Get Two Circulation-Ready White Papers

Two downloadable papers and four upgraded flagship Insights give public-sector, defense, and AI-governance teams more substantial ways to study contestable decisions, changing operational feasibility, human authority, and auditable claims.

August 26, 20265 minute readrelease record
Release
FYI-2026-08-26-02
Availability
Public

Release at a glance

Grid FYI now includes two designed-for-circulation white papers and richer teaching structure across four permanent Insights. The objective is a more useful evaluation packet, not a larger content count: readers should be able to take a paper into a working session, inspect a concrete state model, challenge a fictional case, identify the real authority boundary, and return to the canonical web version for sources and future corrections.

The two portable editions are Accountable Public Decisions and Contested Operations. Both are stable, letter-format editorial artifacts with selectable text, worked tables, diagrams, explicit limits, public-source references, and a link back to their maintained web companion. The web pages remain the canonical accessible reading surface; the PDFs do not claim tagged-PDF or Section 508 conformance.

Two circulation packages

Paper Decision question Read online Download
Accountable Public Decisions How can policy, accepted facts, eligibility, scoring, exceptions, fiscal constraints, authorized determination, notice, challenge, and correction remain connected without transferring public authority to software? Policy That Can Explain Itself PDF
Contested Operations How can facts, assumptions, constraints, authored courses, feasibility, releasability, authority, submission, acknowledgment, and confirmed effect remain distinct as conditions or connectivity change? Course-of-Action Feasibility PDF

These papers are meant to support discussion and evaluation planning. They are not procurement specifications, operational concepts, legal opinions, authority instruments, or assurance reports.

Four flagship Insights now teach the model

Policy That Can Explain Itself now opens with an accountable decision chain and follows the fictional North River case through six preserved revisions. It separates fact correction from exception review, eligibility from advisory scoring, and a funding constraint from both selection and obligation. It also distinguishes a reproducible calculation trace from a contestable explanation that identifies the consequence-bearing fact, source, rule, exception, constraint, authority, and program-defined route for challenge.

Course-of-Action Feasibility now includes a feasibility trace, the complete Course Echo capacity-and-timing arithmetic, and a separate Exercise Northstar event ledger for disconnection and reconciliation. The paper makes a deliberately strict distinction among feasible, releasable, authorized, submitted, acknowledged, and independently confirmed effect. It also links directly to the governed-federation problem in coalition operations.

The Authority Contract for AI-Assisted Decisions now carries an explicit Observe-to-Execute boundary and a state table that distinguishes generation, governed evaluation, recommendation, authorization, and external execution. It separately names system authorization, information-release authority, and operational or program decision authority, because evidence for one does not substitute for the others.

The Auditable Executable Paper now presents a source-to-release claim stack and a compact claim-receipt table. The central boundary remains unchanged: execution can prove that a represented path ran and can preserve the evidence needed to interrogate it. Execution cannot make a source valid, a method sound, an interpretation correct, or a claim true.

The four exhibits share a visual grammar for sources, governed models, human authority, and external effects, but their worked material remains domain-specific. They are not four restatements of the same generic workflow.

How to use the materials

A practical working session can begin with the online paper and its public sources, then use the PDF to annotate one diagram. Replace the synthetic facts with a customer-controlled fixture only after the decision class, source owners, governing rules, unresolved gates, information boundaries, and actual authorities have been named. Write the expected results and stop states before connecting systems or evaluating product behavior.

Keep evaluation planning separate from procurement, deployment, security authorization, information release, and operational or program approval. A successful controlled fixture can justify a bounded product-demonstration claim. It cannot carry the later claims automatically.

Exact availability

Public availability on August 26, 2026 means that these pages and PDF files ship together in the Grid FYI build at their permanent routes. The two PDFs also appear in the Resources library. Consequential future revisions should update the canonical Insight metadata and receive a new release record rather than silently rewriting what this entry says was published.

Evidence and limitations

Official and public sources establish selected institutional context; they do not endorse Grid. The diagrams, tables, and fictional cases are Grid-authored synthesis, not customer cases or operational records. This release does not claim that the current Grid runtime implements every depicted workflow.

Nothing in these materials establishes legal advice, benefit or grant eligibility, due-process sufficiency, acquisition approval, policy or doctrine conformance, authorization to operate, security accreditation, classified or CUI handling approval, coalition releasability, interoperability, model validation, operational authorization, mission effect, performance, cost savings, compliance, or measured outcome. The responsible institution must determine the governing law, policy, doctrine, roles, data rights, controls, integrations, testing, acceptance, release, and authorization for its actual use.

How to verify this release

Open both PDF routes and confirm that they are distinct multi-page artifacts. Open all four Insight pages and confirm the new visual exhibit and worked material. Follow the internal cross-links, and inspect the evidence label, review metadata, and limitations on each page. Treat Public as an editorial availability statement, never as evidence of product deployment or customer use.

Related films, scenarios, and next steps

Choose the next move

Test the claim with a different kind of evidence.

For public-sector leadersTake the working resource into the conversationUse the printable assessment or brief to make assumptions, authority, and remaining proof concrete.See the modelAI with Human AuthorityFive AI-written versions of the same requirement return different answers; the film moves the rule into one visible model and keeps approval with a person.For technical evaluatorsFollow the concept into Grid DevelopersContinue into the linked Grid Developers guide for the exact behavior, prerequisites, and limits used by this explanation.

Continue exploring

Follow the next question.