Features · AGI Future Foundation
The Open Door — Open Global Investment and the Race Nobody Has to Win
There is a particular kind of fear that has shaped every great-power technology race in living memory: the fear of being left out. A country that believes it will be excluded from the next great engine of wealth does not sit still and accept it. It races. It spies. It sabotages. It deploys before it is ready, because deploying early feels safer than being left behind. Open Global Investment is the AGI Future Foundation's proposed answer to that fear, and its logic is deliberately counterintuitive. Instead of guarding the upside, open it. Instead of excluding rivals, invite them in. Give them a stake in the safe success of the enterprise, so that undermining it would mean undermining themselves. It is the most ambitious idea in the family's papers — and, as of today, it has no terms, no participants, and no track record.
The race nobody wins
The pitch deck that carries Open Global Investment states the problem in the plainest possible terms. AGI kept as a closely guarded national asset gives every capable party the same incentive: race to build it first, keep others out, and accept safety risks rather than fall behind. Unilateral development incentivises espionage, sabotage, and rushed, unsafe deployment, because each side fears exclusion and dependency. A programme benefiting only one country concentrates the transition's wealth in one place and leaves everyone else with a reason to undermine it. That is the status quo the model is built against: high incentive to race driven by fear of exclusion, espionage and sabotage, rushed deployment, and a high risk of existential conflict. It is a description of a race nobody wins, because the prize for winning it is a technology built too fast to be safe, owned by one side, and resented by every other.
Open Global Investment proposes to change the incentives rather than the players. The mechanism is a separation ordinary ownership does not make. In a conventional company, the people who own the most economics usually control the most votes, so buying a large economic stake buys influence. Open Global Investment proposes to break that link: upside distributed as widely as possible, while control is concentrated in high-vote shares held by stewards and the Foundation — democratised profit, centralised safety. The pitch standard calls the result a fairer global distribution, and it illustrates the claim with two Gini figures, roughly 0.96 for a United States-only nationalised programme and roughly 0.89 for the open model. Those figures are pitch illustrations, not sourced statistics, and they must not be quoted to an investor, donor, or regulator as measured data. If they are used at all, they must be labelled as a pitch illustration, not a measured forecast. The idea they illustrate, though, is the real subject: that who gets the money and who makes the decisions need not be the same people.
The strategic logic behind the open door is the part of the design that sounds most like a paradox and is most worth taking seriously. By allowing foreign nations and international rivals to participate in the profits of the 33 W3 LLC companies, the pitch argues, the family fundamentally alters their geopolitical calculus. If an adversary is heavily invested in Class A and Class B shares, destroying or recklessly racing the Foundation transforms from an act of national security into an act of profound economic self-destruction. The pitch calls this co-opting rivals instead of racing them, and its slide on the subject promises low incentive to race, a vested interest in the Foundation's secure and safe success, and a cooperative, stabilised transition to post-scarcity. It is unproven, with no terms, no participants, and no track record today. It is also the only idea in the family's papers that takes the race seriously as a problem of incentives rather than a problem of engineering.
There is a limit to the argument, and the papers state it. Open Global Investment is one of three interlocking gears, alongside the Fiduciary Shield and the Windfall Clause, and it does not work alone. The Shield is the legal protection for the safety mission. The Windfall Clause is the distributive engine for global wealth. Open Global Investment is the economic inclusion that defuses conflict. Without the Shield, the open door would be an invitation to buy control. Without the Windfall Clause, it would be an invitation to buy a share of profits that never reach humanity. The three gears are meant to turn together, and the pitch calls them the only viable bridge to a cooperative post-AGI future. That is a large claim for a mechanism with no terms. It is also the claim the rest of this feature has to test, gear by gear, starting with the shares themselves.
Class A: the world buys in
Class A is the broad-access class, and its design is the simplest of the three: one vote per share, full economic participation, optimised for global public access and widespread wealth distribution. The pitch diagram shows it as almost entirely profit participation, with only a small sliver of voting power. That picture is the point. Class A holders own economics, not safety decisions. A Class A majority alone cannot amend the Windfall Clause or override Class C. Class A holders are told upfront what they are buying into: economics, not control. That disclosure is not a footnote. It is the term that makes the rest of the structure honest.
Class A is the class a SAFE in the proposed PBC track would convert into, and it is the class through which the pitch's promise of broad ownership is meant to be kept. Widely open to international investment, drastically reducing expropriation risk — that is the pitch's language for what Class A is for. But widely open does not mean open to anyone, anywhere, without securities or sanctions screening. Class A is open globally, subject to securities law per jurisdiction — Regulation D in the United States, with no general solicitation unless Regulation D Rule 506(c) and verification are chosen. The openness is real, and it is bounded by law, and the bound is part of the design rather than an exception to it.
The economics of Class A are full but not unlimited. Class A fully participates in economic upside, up to the Windfall threshold. Below the threshold, profits follow the normal rules: standard returns to investors and retained funding for research and development. Above the threshold, the excess becomes Windfall Profits, paid to a Windfall Administrator for distribution to four benefit buckets — Global Public Goods, Healthcare Access, Energy Infrastructure, and Technological Self-Determination — for projects advancing the UN Sustainable Development Goals. Class A holders therefore buy into a structure where the upside is real and the ceiling is deliberate: the Windfall Clause is a pre-commitment to share profits above a threshold with humanity, and it reduces the profits available to shareholders above the threshold, by design. Investors must see it as a cost and a risk, not as a benefit, unless counsel drafts it as a participant benefit expressly.
There is a sentence in the pitch materials that must be read carefully, because it is easy to misread in the direction of a promise. The pitch describes Open Global Investment as offering guaranteed participation in the economic upside. Guaranteed participation means participation in upside if there is upside. It does not mean a guaranteed return. No participant, in any class, gets a guarantee of liquidity, a secondary market, or a buyback. Legal documents must say that expressly. A reader who confuses participation with a promise has misunderstood the only class that is open to the world — and the misunderstanding would be the family's fault if the documents did not correct it. They do correct it, in the pitch standards, in the raise structure, and on the governance pages. This feature corrects it again here, because it is the easiest point in the structure to get wrong.
Class A is also the class that carries the model's claim to fairness. Everyone can own the future, but the future's safety cannot be subjected to a hostile takeover — that is the pitch's summary of the separation, and Class A is the everyone in that sentence. The claim is that a widely distributed economic stake is fairer than a concentrated one, and that a structure which lets the whole world own the economics while the Foundation holds the control is the way to keep both promises at once. Whether that claim survives contact with real securities law, real sanctions screening, and real tax advice in dozens of jurisdictions is an open question, and the papers do not pretend otherwise. The eligibility rules, the price, and the governance rights for Class A are all blank. What is not blank is the intention: the broadest class, the fullest economics, and the fewest votes. That is the door, held open.
Class B: the restricted room
Class B is the restricted class, and its restriction is the point. Ten votes per share, economic participation plus voting — roughly half each in the pitch diagram, which shows Class B as both economic and voting, not governance-only. Reserved for aligned partners and sovereign entities meeting strict safety and geopolitical frameworks. That sentence has three load-bearing words: restricted, aligned, and frameworks. Restricted means Class B is not open to anyone who can pay. Aligned means the holder must meet criteria the family has not yet written. Frameworks means those criteria must exist as a real screening process, not as a feeling about who seems friendly. A sovereign that does not meet the safety and geopolitical framework does not get Class B, and does not get Class B voting power by buying Class A.
The distinction between Class A and Class B is the distinction the whole model turns on, and it is worth stating as plainly as possible. Class A is the broad-access class. Class B is the restricted class. Keep that distinction in every document. Class A is for the global public. Class B is for aligned partners and sovereign entities that have passed a framework the papers have not yet defined. Buying Class A does not buy Class B votes. Buying Class B does not happen without screening. Any sovereign or rival participation needs, in addition to the offering documents, sanctions screening, export-control and national-security review where applicable, tax advice, and securities advice per jurisdiction. The pitch's invitation to rivals is therefore not an open invitation. It is an invitation to a room with a door, a framework, and a bouncer — and the bouncer has not been hired yet.
That screening matters because Class B carries real voting power. Ten votes per share is ten times the voice of Class A, and a Class B holder who is also an economic participant has both a financial interest and a governance interest in the enterprise. The pitch diagram shows Class B as roughly half economic participation and half voting power, which is a way of saying that Class B holders are partners, not merely investors and not merely governors. They share in the upside and they share in the decisions — within the limits the Foundation's Class C control sets, and within the framework that admitted them. Transfer is restricted too: transfer only to a qualified holder meeting the framework, or conversion to Class A on transfer to a non-qualified holder, with counsel to set the exact rule. A Class B share that could be sold to anyone would not be a restricted share for long.
The framework itself is the largest blank in Class B, and the papers are candid about that. The safety and geopolitical criteria, the screening process, the qualified-holder definition — all are listed as terms counsel must set before any Open Global Investment solicitation. There is no list of aligned partners. There is no list of sovereign entities. There is no description of what alignment means, beyond the phrase strict safety and geopolitical frameworks. That blank is not a detail. It is the difference between a model that invites the right rivals and a model that invites the wrong ones, or invites none at all because nobody can pass a test that has not been written. Until the framework exists, Class B is a room with no admission criteria, and a room with no admission criteria is either empty or full of the wrong people.
There is also a question the papers raise and do not answer: what happens when a Class B holder stops being aligned. A sovereign that met the framework in one year may not meet it in the next. A partner that was aligned when it bought in may diverge. The transfer and conversion rules are meant to handle that — conversion to Class A on transfer to a non-qualified holder — but conversion on transfer is not the same as conversion on divergence, and the papers do not say which one applies. Counsel must set it, and until counsel does, the honest statement is that Class B is a restricted class with an unwritten restriction. The door to the restricted room exists. The criteria for entering it do not.
Class C: control without economics
Class C is the control class, and it is the strangest of the three, because it owns nothing and controls everything. One thousand votes per share, held exclusively by the Foundation, with no economic participation at all — control only. The pitch diagram shows Class C as voting power only, with no economic gold at all. Even a small number of Class C shares dominates Class A at one vote and Class B at ten votes on any ordinary vote. That is the pitch's mechanism for absolute, unshakeable ethical control, and for the claim that safety cannot be subjected to a hostile takeover. It is also the mechanism that makes the other two classes honest: Class A and Class B holders own economics, and they do not control safety decisions, because Class C does.
The Foundation's control over the 33 W3 series is exercised through three channels, and the papers list all three: Class C control at the PBC level, mission and voting agreements between the Foundation and each series, and each series' Steward Council mission gate. Control, the papers add, is not ownership. The Foundation does not own the series' profits by virtue of controlling them. Control and ownership are different things, and confusing them would mislead everyone: the Class A and Class B holders who own the economics, the series managers who run the companies, and the public that is asked to trust the structure. Class C is control without economics, and that separation is the whole point of the class. A control share that also took economics would be a different instrument, and a worse one: it would give the Foundation a financial interest in the decisions it is meant to police.
Class C is non-transferable except to a successor Foundation board or trustee as the Articles provide, and counsel must set the conversion and cancellation trigger, so control does not become a perpetual unaccountable block. That last clause is easy to skip and important to keep. A control class that could never be changed, never be transferred, and never be cancelled would be a permanent lock on the enterprise, held by whoever happened to control the Foundation at the moment the lock was set. The papers require a trigger — a defined event on which Class C converts or is cancelled — precisely to prevent that. The trigger is blank, for counsel to set. But the requirement that there be one is not blank, and it is the difference between a control mechanism and a dynasty.
There is a disclosure duty that runs with Class C, and the papers state it in several places. Class C's thousand-vote control must be disclosed to Class A and Class B investors as a governance risk — they own economics, they do not control safety decisions — not hidden behind decoupled language. The pitch's phrase democratised profit, centralised safety is a summary, not a disclosure. A disclosure says: the Foundation holds shares with a thousand votes each, those shares carry no economics, they dominate ordinary votes, and they can block or direct safety decisions regardless of what the economic holders want. An investor who buys Class A or Class B without understanding that has not been told the truth about what they own. The papers require the truth to be told. This feature tells it again.
Class C is also the class that makes the Windfall Clause enforceable, and that connection is worth making explicit. The clause may be amended, suspended, or repealed only by a supermajority vote plus the Class C vote, and a purported amendment that reduces the Windfall Share, raises the threshold, or removes a benefit bucket, adopted without that vote, is void. Class C's thousand-vote control is part of that protection — and Class C holders must disclose that control to Class A and Class B investors as a governance fact, not as a guarantee that the clause will never change. A protection that depends on a vote is not a guarantee. It is a hurdle, and a high one, and the papers are careful to describe it as a hurdle rather than a wall. That care is the difference between a promise and a mechanism.
The blanks: every term that does not exist
Open Global Investment has no terms today, and the papers list the blanks with a completeness that is itself a kind of honesty. Participation instrument: Class A shares, profit participation interests, revenue share, or something else — choose one, with a securities analysis for that instrument. A participation interest that is, in substance, a security needs an offering document, even if it is called participation. Eligibility: who may participate — accredited only, international participants, rival and sovereign participants as the site suggests — with each category needing securities, sanctions, tax, and anti-money-laundering advice. Price and valuation: set at the time of the participation offering, not today. Governance rights: none, information rights only, or an advisory vote — state expressly, with a default of none unless counsel drafts a right. Transfer and liquidity: none until a liquidity event, limited transfer with board approval, or something else — state expressly, and do not promise liquidity, a secondary market, or a buyback. Reporting: annual benefit and financial reporting to participants, as counsel advises.
That list is long, and it is not the whole list. The Windfall Clause, which is the other half of the open-door promise, has its own blanks, and they are larger. The threshold is expressed as a percentage of Global GDP, not a fixed dollar amount — but the percentage itself is blank, the GDP measure is blank, the profit measure is blank, and the scope is blank. The Windfall Share is blank: the pitch calls it a massive portion, but gives no number, and none may be quoted as the share. The Administrator is blank. The auditor, the dispute forum, the payment timing, and whether the clause survives a sale, merger, or change of control are all blank, for counsel to confirm in writing. A percentage of Global GDP with no percentage, no measure, and no scope is a direction of travel, not a trigger. Two guardrails are built in: the clause must be adopted before a series becomes profitable, and it may be amended only by a supermajority vote plus the Class C vote. And the clause must appear in each series' designation documents, not only in a parent company's articles — today it has no series to bind.
The sequencing rules are the papers' answer to the blanks, and they are strict. Do not open Tranche 3 — Open Global Investment — until Tranche 2 terms, the benefit report, and counsel's Open Global Investment terms are all signed. Tranche 3 is not a SAFE, not a priced round, and not a donation. It is a separate participation programme that needs its own documents. Do not combine Open Global Investment with a SAFE in one document, or promise a SAFE holder automatic Open Global Investment participation, unless counsel drafts that combination expressly. Do not put a Windfall Clause in a SAFE or equity offering document as a promised return to investors — it is a commitment to share profits away from shareholders, so investors must see it as a cost and a risk, not as a benefit, unless counsel drafts it as a participant benefit expressly. The blanks are not an invitation to fill them in with enthusiasm. They are an instruction to wait for counsel.
There is a temptation, when a design has this many blanks, to treat the blanks as a reason to dismiss it. That would be a mistake, and the papers explain why. A pre-commitment has to be made before the profits exist, because it cannot be made after. A board that means to share profits later can be sued, outvoted, acquired, or worn down before the profits arrive. The Windfall Clause answers that by making the sharing decision before the profits exist — hardwired into the governing documents, so sharing is the default. The blanks do not change that logic. They change what can be promised today, which is nothing, and what must be done next, which is counsel's work. The design is a direction of travel with a mechanism for getting there. It is not a destination that has been reached.
And that is the honest place to leave the blanks: visible, listed, and unfilled. The participation instrument is blank. Eligibility is blank. Price and valuation are blank. Governance rights are blank, with a default of none. Transfer and liquidity are blank, with no liquidity, secondary market, or buyback promised. The threshold percentage is blank. The GDP measure is blank. The profit measure is blank. The scope is blank. The Windfall Share is blank. The Administrator is blank. The auditor is blank. The dispute forum is blank. The payment timing is blank. Survival through a sale or merger is blank. A feature that filled any of those blanks with a guess would be inventing the terms it claims to describe. This one leaves them blank, because the papers leave them blank, and the papers are the source.
The fork: which issuer, which track
There is a fork underneath Open Global Investment, and it decides everything about who can offer it, to whom, and under what law. As filed and verified, the Foundation is a California Nonprofit Public Benefit Corporation, Entity No. B20260196664, Articles filed 2026-04-24, with Secretary of State approval 2026-04-30, principal and mailing address 995 Market St AGI Lab -10, San Francisco, CA 94103, agent Mike Rice, and an EIN assigned 2026-05-07. A nonprofit with this filing cannot sell equity, SAFEs, profit shares, or tokens, and cannot distribute profits. As labelled on the live website, PBC is used in the for-profit sense: a benefit corporation with dual-class equity, Class A profit participation, and Open Global Investment. Those are not the same entity type, and the corporate raise advice follows the filed status, because a court, the Attorney General, the IRS, and an investor's counsel will follow the filing, not the website label. The fork is unresolved, and until counsel chooses one track in writing, no solicitation should use the for-profit figures.
The three paths the papers lay out are distinct, and each one changes the open door. Form a new for-profit California benefit corporation, with its own filing, EIN, and bank account, never reusing the nonprofit's entity number or EIN, and run the raise through that new entity, with the nonprofit kept separate. Convert or restructure the existing entity — a major Attorney General and IRS transaction, not a name change, not to be assumed available or tax-free. Or run the raise through AGI Corp instead, keeping the Foundation as a nonprofit mission steward. Until one path is chosen in writing, the Open Global Investment content on this page must not be attached to the nonprofit Foundation or used to solicit investment in the PBC. An investor would not know which entity they were buying into, and that uncertainty is itself a securities problem.
The fork also decides the relationship between the PBC and the nonprofit Foundation, and that relationship is delicate. If the nonprofit Foundation remains separate, the PBC's benefit purpose should align with, but not be controlled by, the nonprofit, unless a written mission agreement expressly provides a mission veto — and counsel must advise whether a nonprofit veto over a for-profit PBC creates a control or private-benefit issue. The Foundation's control over the series, through Class C and through mission agreements, is a control over the commercial family that the nonprofit must exercise without becoming its owner, and without spending its donations on its commercial purposes. The open door, for all its talk of global inclusion, is therefore also a test of the Foundation's independence: can it hold the control class, enforce the mission gate, and administer the Windfall, without becoming a captive of the enterprise it governs.
The sequencing the papers set out is the discipline that keeps the fork from becoming a solicitation problem. Choose the issuer in writing first. File or confirm the PBC, with its benefit purpose in the Articles, its good standing, its EIN matching its SOS name, its bylaws, its board, and its Mercury account in its SOS name. Clean up the site, reconciling the Foundation site — nonprofit apex, 33 subsidiaries, Invest section — with the PBC raise story, so there is one issuer, one count once series are designated, and one solicitation channel in private offering documents, not public site buttons. Authorise the board, with one cap, one discount, and one round size, plus a disclosure document and risk factors including benefit-purpose risk, dual-class control risk, Windfall cost risk, and multi-entity risk. Close Tranche 1 with accredited investors only, with a SAFE register and a Form D within fifteen days. Reach the milestone gate before Tranche 2. And open Tranche 3 — Open Global Investment — and the Windfall terms only after counsel drafts signed terms, never as a promise inside Tranche 1 or Tranche 2 documents.
That is where the open door stands: a design with a clear logic, a clear share structure, a clear set of blanks, and a fork that has not been resolved. The logic is that rivals with a stake have less reason to race. The structure is three classes: one vote and full economics for the world, ten votes and economics plus voting for the restricted few, a thousand votes and no economics for the Foundation alone. The blanks are every operative term. The fork is which issuer will carry them. Nothing has been offered, nothing has been sold, and no participant has been admitted, because there is no issuer, no framework, and no terms to admit them under. The door is drawn on the plans. It has not been built, and it has not been opened. The question the design poses — whether openness can defuse a race — remains unanswered, and it will stay unanswered until the terms exist, the issuer is chosen, and the first participant walks through.
What this is / What it isn’t
What this is
- A full deep dive on Open Global Investment: the race-to-the-bottom logic, the three-class share structure, the screening that Class B requires, and the control that Class C holds without economics.
- A story about incentives rather than trust: rivals are invited into the upside so that undermining the enterprise would mean undermining their own stake.
- A public record of what is blank: every operative term — instrument, eligibility, price, governance, transfer, threshold, share, administrator — left for counsel to set.
What it isn’t
- Not an offering: no participation interest has been issued, no participant has been admitted, and nothing here is a price, a term, or a promise of return.
- Not a guarantee: guaranteed participation means participation in upside if there is upside, not a guaranteed return, and no liquidity, secondary market, or buyback is promised.
- Not the filed Foundation's programme: the nonprofit Foundation cannot issue shares, SAFEs, profit shares, or Open Global Investment interests, and the for-profit PBC track is proposed, not formed, with the issuer not yet chosen.
Where to see it
The governance pages carry the full detail behind this feature: Open Global Investment, Share Structure, Windfall Clause, and Fiduciary Shield, with the Steward Council holding the mission gate. The FiduciaryCorporateShield repository is public on GitHub: FiduciaryCorporateShield — a two-line README, title only, not a built system.
Nothing on this page is legal, tax, or financial advice. Open Global Investment, the three-class structure, and the Windfall Clause are draft designs for a for-profit benefit-corporation track that counsel has not yet chosen; the filed entity is a California nonprofit public benefit corporation, and no IRS 501(c)(3) determination letter is in the files. Gini figures are pitch illustrations, not measured data.