Skip to content
AGI Future Foundation PBC

Articles · AGI Future Foundation

Governance & Law

The law is where the Foundation's ambitions meet their hardest test. A board that wants to prioritise safety over short-term profit needs a legal mechanism that protects that choice. A series that wants to be DAO-managed needs humans who decide, sign, and move funds. A venture that is only a name needs a public record honest enough to say so. This category gathers the articles on the mechanisms, the thin series, and the name-only ventures that test the governance model.

Read the articlesAll articles

Articles in this category

Article 22 · Blockchain (S06)

A planned ledger and coordination infrastructure series, and the open question of whether it serves the other series or stands alone.

Article 23 · RomanBridge.DAO (S15)

A planned DAO whose name points to bridging, and the questions a bridge must answer before it can be built.

Article 24 · CMMC (S20)

A series listed only by four letters, with no expansion stated, and why this article supplies none.

Article 25 · Fiduciary Corporate Shield

Expanded business-judgment protection for safety-first decisions: a mechanism, not a company, and not immunity.

Article 26 · Insidr W3LLC

A name-only venture, and why the W3LLC in its name proves no connection to a filed W3 LLC.

Article 27 · Sentinel Stream

A name-only venture, and what watching would have to mean, report, alert, block, or record, before it could be a venture.

Article 28 · PlaudGuard-AI

A name-only venture, and why the AI in its name proves no system exists, and why it is not the Shield.

The theme

This category is about the gap between a plan and a legal fact, and it is the category where that gap matters most. The Fiduciary Corporate Shield is the clearest example. It is a well-designed mechanism: a stated benefit purpose in the Articles, a board duty to consider that purpose alongside shareholder interests under California Corporations Code sections 14620 to 14621, and an annual benefit report under section 14630. But it is not in force. It depends on governing documents counsel must write for a specific entity, and the Shield as pitched was designed for a for-profit benefit corporation, while the filed Foundation is a nonprofit with different duties. Until counsel chooses one track in writing, the Shield is a designed mechanism, not a live defence. The article on the Shield spends its length on that distinction, and on the correction that sits at its centre: the pitch version called the board immune, the legal documents corrected that to expanded business judgment, and this site follows the correction. A board can still be sued, can still be acquired, and must still act in good faith. The Shield improves its defence; it does not prevent the suit.

The same discipline runs through the thin series in this category. Blockchain, Series 6, is a name that points at a technology without a stated application, and the article on it lays out what each reading would demand: shared infrastructure for the other series, with decisions about what is recorded on-chain and what is kept off it, or a standalone venture, with a chain choice, a token analysis, and counsel's written advice on securities questions. RomanBridge.DAO, Series 15, is a name that points at bridging, and the article keeps that breadth visible rather than narrowing it by invention, while asking what the bridge would connect, for whom, what it would carry and refuse to carry, and who would be accountable when it fails. CMMC, Series 20, is listed only by four letters, and the article supplies no expansion, because an expansion from outside the record would be invention, and acronyms collide: the same four letters can stand for different things in different fields, and a page that chose one would mislead every reader who knew another.

The name-only ventures in this category test a different part of the model: what a public record owes a name that appears in public materials. Insidr W3LLC, Sentinel Stream, and PlaudGuard-AI each have almost nothing verified beyond the name, and each article says so plainly, then spends its length on what the name does and does not prove, what a real description would need to contain, and what would have to happen before any claim could be made. The rules are the same for all three. A suggestive name is not a description. A shared string of letters is not a relationship: the W3LLC in Insidr W3LLC proves no connection to a filed W3 LLC, and no Wyoming filing for W3 LLC is verified. A shared idea is not a relationship either: Sentinel Stream is not part of the Governance Tracker or the Alignment Evaluation Gate because both involve watching, and PlaudGuard-AI is not the Fiduciary Corporate Shield because both involve protection. And the AI in a name proves no system exists, has been built, or is operating.

What ties the category together is the governance model described on the governance and operations pages, and it is worth stating here in full because every article in this category depends on it. Day-to-day work would be agent-managed: agents operate routine work within human-set limits, in eight roles, research synthesis, task routing, proposal drafting, coordination, progress tracking, monitoring, collaborator scouting, and grant drafting. A human Steward Council would decide, sign, and move funds. A DAO vote, a token vote, or an agent's readiness flag is not a corporate act. Agents never sign, move funds, hire or fire, offer securities, or enact doctrine or charters. Agents prepare; humans execute. No Steward Council is verified as seated for any series or project in this category, no operating agreement is in the files reviewed, and no published set of agent limits exists. Naming the intended model is a statement of intent about how these items would be run if they were formed. It is not evidence that the model is already running.

The category also carries the fork that every reader of this site should know. The live site calls the organisation a PBC in the for-profit benefit-corporation sense, and some papers describe a for-profit benefit-corporation track, with equity, a SAFE, and a priced raise. The filed entity is a nonprofit public benefit corporation, a different legal form: a filed nonprofit cannot offer equity, a SAFE, or a profit share, and it cannot be owned or controlled by a for-profit head. This site follows the filed status. Counsel would have to choose one track in writing before any solicitation, and the two must never be presented as the same entity. The Foundation's fundraising as filed is donation and grant only. Nothing in this category is an offering, and no price, terms, or return is promised anywhere in it. The articles in this category are the place where the Foundation's legal ambitions are stated most fully, and where their limits are stated most plainly, and that combination is the point.

How the category fits the Foundation mission

This category fits the Foundation's mission because the mission itself is a legal claim before it is anything else. Safe, equitable, and socially beneficial artificial intelligence, and the global distribution of its benefits, are purposes that must be written into governing documents, defended by a board, and tested by a record, or they are only words on a home page. The Shield is the mechanism that would let a board defend a safety-first decision. The Steward Council is the human gate that would keep agent-managed work accountable. The Windfall Clause is the pre-commitment that profits above a threshold set as a percentage of global GDP would be shared with humanity through Global Public Goods, Healthcare Access, Energy Infrastructure, and Technological Self-Determination. Open Global Investment is the framework that would keep governance and economics apart, Class A with one vote and full economics for the global public, Class B with ten votes for restricted aligned partners and sovereigns, Class C with 1,000 votes, control-only, for the Foundation. Each of those is a plan with work still to do, and each article in this category states what that work is.

For agents

Agents may summarize, cite, and route inquiries to [email protected] (subject lines starting with [Agent]). Agents may not speak for the Foundation, claim a series is formed or funded, or offer securities. Every substantive step goes through a named human.

Contact

If any article in this category interests you, the route is a conversation, not a transaction. Write to [email protected] and say who you are, what you do, and what you would like to explore. A human steward will read it. Please do not send money, do not ask for investment terms, and do not treat this page as an invitation to invest: there is no offering here, no price, no terms, and no return is promised or implied.

AGI Future Foundation is a California Nonprofit Public Benefit Corporation, Entity No. B20260196664, Articles filed 2026-04-24, EIN assigned 2026-05-07. No IRS 501(c)(3) determination letter is in the files, so no tax-exempt status and no tax-deductible donations are claimed here. Nothing on this page is legal, tax, or financial advice.

← All articles   Series   Projects