We The People Amendment
The Reasoning
Every provision of this amendment, one at a time. What problem it solves, how it solves it, what changes, what happens if nothing is done, what it rests on in law and history, and who has an interest in defeating it.
Each provision in the amendment links directly here. You can read one and ignore the rest.
Contents
Section 1
Section 2
Section 3
Section 4
Section 5
Section 6
Section 1
Rights belong to human beings
The problem
The Constitution never defined the word person. It uses it constantly and leaves it open. For most of American history that was harmless. It stopped being harmless once courts began treating artificial entities as persons for constitutional purposes, because the Fourteenth Amendment protects “any person” and a corporation that counts as a person inherits protections written for human beings.
The origin is worth knowing because it is not what most people assume. In Santa Clara County v. Southern Pacific Railroad (1886) the Supreme Court decided a tax case and said nothing about corporate personhood in its opinion. The claim that corporations are persons under the Fourteenth Amendment appears in the headnote — a summary written by the court reporter, not by any justice. Later courts cited the headnote. A summary became a doctrine.
How this provision solves it
The sentence is affirmative rather than prohibitive. It does not say entities are denied rights; it says the rights belong to natural persons. A prohibition invites courts to find exceptions. An assignment leaves nothing to except, because the rights already have an owner.
The definition works by exclusion. The operative word is individual: an individual cannot be a group, a class, a corporation, or a system. The text never states when a human life begins, and that omission is deliberate and load-bearing.
Net effect, and the cost of doing nothing
Effect: every constitutional argument that begins “this corporation has a right to…” loses its foundation. Regulation of artificial entities becomes an ordinary legislative question decided by voters, not a constitutional one decided by judges.
Cost of inaction: the doctrine is not static. It has expanded through free speech, religious exercise, and property claims, each extension citing the last. There is no natural stopping point, and no one ever voted for any of it.
A person who can be commanded is not a person
There is a simpler objection to corporate personhood than any doctrinal one, and it is worth stating in plain terms.
A corporation does not decide anything. Named human beings decide, under authority the state granted them in a charter. When a corporation “speaks,” a particular person chose the words. That is not a second person exercising conscience alongside the first — it is one person’s voice, amplified by an instrument the state placed in their hands.
A citizen has no right to direct how another person votes or speaks. Yet the effect of corporate personhood is that certain citizens direct the political speech of a legal “person” that cannot refuse them. The chief executive holds their own voice, and command of another. Every other citizen holds one.
The second half of the objection is harder to answer than the first. Directors owe a fiduciary duty to shareholders. A corporation is not merely permitted to subordinate conscience to profit; in the ordinary case it is legally required to. The law therefore created an entity forbidden from acting on conscience, and then extended to it the rights of conscience. Those two propositions cannot both stand.
This is why the amendment addresses the question by definition rather than by argument. It does not weigh whether corporate speech is valuable. It states that a group is not an individual, and leaves the rest to follow.
What it rests on
The Constitution defines almost nothing — not speech, press, religion, cruel, unusual, due process, or commerce. The one conspicuous exception is treason, defined narrowly in Article III with a two-witness requirement, because English monarchs had stretched the word to destroy political opponents. Definition is the founders’ remedy for a documented history of an elastic term being abused. That record now exists for person.
Note also that Article I, Section 2 called enslaved people “Persons” while counting them as three-fifths for apportionment. The framers negotiated around the word rather than redefining it — evidence that in the founding text, person meant human being.
Who will work to defeat it
Opposition comes from any interest whose legal strategy depends on an entity holding rights of its own: organized business federations, large-scale litigation defense practice, and increasingly the artificial intelligence industry, for which this forecloses an entire category of future argument before it can be made.
Section 1
No entity holds constitutional rights
The problem
The corporate question took roughly a century and a half to reach its present state. The equivalent question for artificial intelligence is open right now and nothing in American law forecloses it.
The realistic path is not a machine demanding the vote. It is commercial and incremental: an autonomous system that holds assets, enters contracts, and accumulates the attributes of legal personality through statute and practice until the label follows the substance. That is approximately how corporate personhood arrived, and it arrived without a public decision.
How this provision solves it
The list is deliberately over-inclusive, and the closing phrase — of whatever form now existing or hereafter devised — is the futureproofing. It is standard drafting language for technology and it means a court in 2080 cannot hold that some entity form nobody imagined in 2026 escapes the section.
Recognition by a foreign state is closed off as well, in the companion clause in Section 3. Without that, personhood conferred abroad creates pressure for recognition here through comity and treaty.
Net effect, and the cost of doing nothing
Effect: a permanent bar, established before the question is contested rather than after.
Cost of inaction: legal personality, once recognized, is extraordinarily difficult to withdraw, because the moment it exists the entity holding it can litigate to keep it. Preventing recognition takes one amendment. Reversing it would take an amendment plus the unwinding of everything built on top of it in the interval. That asymmetry is the whole argument for acting early.
What it rests on
Several constitutions drafted after 1945 address explicitly whether fundamental rights extend to legal persons. Germany’s Basic Law, Article 19(3), extends them to domestic juridical persons only insofar as the nature of the right permits. South Africa’s Constitution takes a similar calibrated approach in Section 8(4). Both were written by people who had watched what happens when the question is left to accumulate.
This amendment takes the stricter line: no constitutional rights at all, statutory privileges instead. That is the most consequential judgment in the section and it is made deliberately. A corporation is a creature of law. We created it; we can legislate its protections. A sole proprietorship remains an extension of an individual and keeps that individual’s rights.
Who will work to defeat it
Artificial intelligence developers and deployers, most directly. The cost of foreclosing an entire future legal argument is impossible to calculate and therefore easy to fear.
Section 1
Entities may still defend the people behind them
The problem
Every proposal of this kind faces the same attack, and it is a fair one: if artificial entities have no constitutional rights, can police search a business without a warrant? Can a factory be seized without compensation? Can a newspaper be shut down, since newspapers are corporations? Can the NAACP protect its membership list?
That last question is not hypothetical. In NAACP v. Alabama (1958) the state demanded the organization’s membership rolls. The organization prevailed — not because it held rights of its own, but because it was permitted to assert the associational rights of its members, who would have faced retaliation.
How this provision solves it
This paragraph does two things in one sentence.
It states the source of corporate existence. Entities exist by leave of the People and hold only privileges conferred by law. Not rights. Privileges, granted, revocable.
It preserves derivative standing. An entity may appear in court to assert the rights of the humans it lawfully represents. That is exactly the NAACP v. Alabama mechanism, written into the Constitution rather than left to doctrine.
Net effect, and the cost of doing nothing
Effect: a civil rights organization can still defend its members. A union can still act for its workers. A business can still assert its owners’ rights against an unlawful search. What disappears is the entity claiming a right of its own, independent of any human being.
Cost of inaction: without this paragraph the section would be indefensible in its first hearing, and it would deserve to be.
What it rests on
The distinction between holding a right and asserting another’s is ordinary in American law. Guardians, executors, trustees, and class representatives all act for people whose rights are at stake without owning those rights themselves. This paragraph applies a familiar structure to a new question.
Who will work to defeat it
Civil liberties organizations may still object, and their objection is the most important one to answer well. They litigate as corporations and have a genuine, principled concern that narrowing entity rights weakens their ability to bring cases at all.
That concern is legitimate and this paragraph is drafted for it. If it does not satisfy them, the defect is in the drafting and should be fixed rather than argued around. These are the first people this section should be shown to.
Section 1
Hiding the humans is fraud
The problem
Political money now moves through structures designed so that no name is ever attached to it. A committee with a patriotic name, a nonprofit that discloses nothing, a chain of entities ending somewhere no voter can follow. The voter sees a message and cannot find a person behind it.
The same technique now appears in synthetic communication: an account, a voice, or a face presented as a human being where no human being exists.
How this provision solves it
The clause reaches three distinct acts: holding out as a person, choosing a name calculated to obscure, and operating so that a reasonable person would be misled.
Then it names the offense. Concealment is fraud — not a reporting violation, not a civil infraction. Fraud.
Net effect, and the cost of doing nothing
Effect: the constitutional footing for attribution law, and the removal of the corporate veil as a shelter for it.
Cost of inaction: every disclosure law can be defeated by adding one more entity to the chain. Rules that attach to form rather than to concealment will always be one reorganization behind.
What it rests on
This is the strongest ground in the section and it should be stated that way. The corporate veil has never protected fraud. Courts have pierced it for fraud for as long as the veil has existed, and corporate officers have always been personally liable for torts they personally commit.
This amendment does not invent an exception to limited liability. It constitutionalizes one that already exists and makes it uniform across the states.
Who will work to defeat it
The professional practice built around corporate structure as shelter, and every organization whose political effectiveness depends on its funders remaining unnamed — a category that spans the political spectrum and will be quieter in its opposition than organized business.
Section 1
What this section does not do
The problem
Three failure modes, each of which would sink the amendment if left unaddressed.
Liability. An earlier draft of this text excluded entities from “rights, privileges, immunities, responsibilities.” Read literally, that eliminated corporate liability — the precise opposite of the intent, and a gift to the corporate defense bar.
Governments. States, counties and cities are artificial entities created by law. Without an exclusion the section would strip California of its own constitutional standing.
Personhood at the edges. Any positive definition of when a human being begins converts this into an abortion amendment and ends it.
How this provision solves it
All three are handled in one sentence, expressly rather than by implication. The third clause is the critical one: neither enlarges nor narrows the class of natural persons under existing law. Whatever the law says the day before ratification, it says the day after.
Net effect, and the cost of doing nothing
Effect: a pro-life legislator and a pro-choice legislator can both vote yes without either conceding anything. Very little in American politics can do that, and it should be said aloud and first, before anyone alleges otherwise.
Cost of inaction: without this clause, every hearing becomes a hearing about abortion, and the corruption provisions are never discussed again.
What it rests on
This is a savings clause, the most common device in constitutional drafting. Compare the Ninth Amendment: the enumeration of certain rights “shall not be construed to deny or disparage others retained by the people.” That does not concede uncertainty about which rights exist. It prevents an inference from silence. This clause does the same work.
Who will work to defeat it
Advocates on both sides of the personhood question who would prefer this amendment as a vehicle for their own. They should be told plainly that it is not one, and that the refusal is what makes ratification possible.
Section 1
Courts are told how to read this
The problem
A rule that only states what is true can be narrowed by a court that reads it grudgingly. Corporate personhood was built from a headnote. It can be rebuilt from an inference.
How this provision solves it
A rule of construction is an instruction addressed to judges rather than to citizens. It does not describe conduct; it forecloses a reading.
The grammar is borrowed deliberately. The Eleventh Amendment begins: “The Judicial power of the United States shall not be construed to extend to…” This is the document’s own language, not an invention, and readers attentive to original forms notice.
Net effect, and the cost of doing nothing
Effect: closes the route by which the doctrine was built in the first place.
Cost of inaction: the substantive rule survives and the inference reappears elsewhere in the document.
What it rests on
Amending the Constitution to reverse the Supreme Court is ordinary rather than radical. The Eleventh Amendment reversed Chisholm v. Georgia. The Fourteenth reversed Dred Scott. The Sixteenth reversed Pollock. The Twenty-Sixth reversed part of Oregon v. Mitchell. Four of twenty-seven ratified amendments exist substantially to overturn specific decisions. Anyone who calls this an attack on the Court is describing Article V.
Who will work to defeat it
Institutional objections about judicial independence, which will be raised sincerely and should be answered with the list above rather than with argument.
Section 2
Money is not speech
The problem
In Buckley v. Valeo (1976) the Supreme Court held that limits on campaign expenditures restrict political speech, because spending money is how speech is amplified. That holding is the foundation of everything that followed: First National Bank of Boston v. Bellotti (1978), Citizens United v. FEC (2010), McCutcheon v. FEC (2014).
The practical result is that the ability to influence an election scales with wealth, and legislatures cannot correct it, because correcting it is treated as censorship.
How this provision solves it
The amendment does not argue with the doctrine. It removes its premise.
Speech is a human act. A payment is a transaction. The first sentence states that flatly, and the second closes the obvious evasions — property, credit, digital tokens, loans, in-kind services. Anything of value.
Net effect, and the cost of doing nothing
Effect: contribution and expenditure limits, public financing, and disclosure requirements stop being First Amendment questions.
Cost of inaction: a candidate’s first duty remains fundraising, and fundraising means pleasing whoever has money to give. That filter operates before any vote is cast and determines who can run at all.
What it rests on
This distinction is not novel. The law routinely separates money from expression: bribery is illegal though it involves a payment and a request; securities law regulates what may be said to sell a stock; commercial speech has always received reduced protection. The proposition that money is categorically speech is the outlier, not the rule.
The Fourteenth Amendment reference
This provision closes by stating that political power derives solely from the equal voice of individual citizens, as secured by the Equal Protection Clause of the Fourteenth Amendment.
Amendments do refer to one another. The Twenty-First repeals the Eighteenth by name. The Fourteenth’s apportionment section supersedes the three-fifths clause without naming it. The reference here is not decorative: it identifies the constitutional value the money-is-not-speech rule serves, and it signals that the equalization principle rejected in Buckley is restored rather than invented.
Who will work to defeat it
Every interest that currently converts money into political outcomes, and a body of First Amendment scholarship that regards Buckley as correctly decided. The second group will argue in good faith and should be engaged on the merits.
Section 2
No entity money in politics, through any route
The problem
A ban on corporate political spending that names only corporations is defeated on the first day by adding a nonprofit to the chain. Corporate treasury money moves to a social welfare organization, which moves it to a political committee, which buys the advertising. Each step is lawful and the original name never appears.
How this provision solves it
The second clause is the operative one and it is what distinguishes this text from earlier proposals. Funds originating, directly or indirectly, with any such entity, through any intermediary however characterized or organized.
The rule follows the money to its source rather than attaching to the form of the spender. A dollar that began in a corporate treasury cannot be spent on politics regardless of how many organizations it passes through.
Note also the scope: elections, ballot measures, and the enactment, defeat, or administration of law. Lobbying is included. An elections-only rule would not touch the mechanism by which industry shapes regulation after the election is over.
Net effect, and the cost of doing nothing
Effect: two consequences worth understanding before endorsing this.
Unions survive. A union is funded by individual members’ dues. That money originates with natural persons. Unions may spend, subject to full attribution.
Trade associations do not. Organizations funded from corporate treasuries are caught, regardless of their tax status. This is the single largest practical effect of the amendment.
Cost of inaction: disclosure and limitation rules remain permanently one reorganization behind the money.
What it rests on
Following value to its source rather than to its nominal holder is the standard technique in anti-money-laundering law, sanctions enforcement, and beneficial-ownership reporting. The drafting problem here is identical and the solution is borrowed from regimes that work.
Who will work to defeat it
The most concentrated and best-funded opposition in the entire amendment. Trade associations exist to do the thing this clause prohibits, and they are experienced, organized, and permanently staffed.
Extending the ban to lobbying also raises a genuine Petition Clause question. The First Amendment protects “the right of the people… to petition the Government for a redress of grievances,” and paid lobbying has been treated as within it. This amendment, being later in time, controls — but expect that to be the doctrinal center of the fight.
Section 2
No contributions at all; public funds only
The problem
Limits invite evasion. Every contribution ceiling in American history has produced a workaround: bundling, party transfers, independent committees, dark money. The regulatory response is always slower than the innovation.
And a limit leaves the underlying dynamic untouched. A candidate who must raise money spends their time with people who have money, and learns which positions make that easier.
How this provision solves it
The rule is absolute rather than graduated. No contributions, from anyone, in any amount. Campaigns are financed exclusively from public funds.
Self-funding is barred as well. Without that clause a wealthy candidate simply opts out of the system and the equality is illusory.
Net effect, and the cost of doing nothing
Effect: no candidate ever needs to ask anyone for money. The entire architecture of donor access — the call time, the fundraiser, the bundler, the obligation — ceases to exist. A representative’s available hours go to constituents because there is nothing else to spend them on.
Political parties become volunteer organizations. They may not receive contributions or purchase placement. They nominate, organize, and speak. This is the largest structural change in the amendment and most readers will not notice it on first pass.
Cost of inaction: the fundraising filter continues to determine who can run before any voter is consulted.
What it rests on
Full public financing is unusual in the United States but not unknown elsewhere. Several European democracies finance parties and campaigns predominantly from public funds, on the theory that elections are a public function and ought to be publicly paid for, like courts or ballots.
Who will work to defeat it
Incumbents with existing donor networks, the professional campaign-finance industry, and both major parties, whose organizational power depends substantially on their role as money conduits.
Section 2
Nobody buys political advertising
The problem
Even with contributions eliminated, wealth still reaches voters by purchasing access to other people’s audiences. Independent expenditure is the channel, and after Citizens United and SpeechNow.org v. FEC (2010) it is effectively unlimited.
How this provision solves it
The prohibition attaches to the transaction, not to the content. Nobody may buy, and nobody may sell, placement.
The line is clean and worth stating precisely. A newspaper publishing an editorial distributes its own content to its own audience; it purchases nothing. A corporation buying a thirty-second spot purchases access to someone else’s audience. One is expression. The other is a payment. This section reaches payments.
“Targeting” and “amplification” are included because purchased reach no longer arrives only as a bought slot.
Net effect, and the cost of doing nothing
Effect: political communication survives entirely. What ends is the ability to buy a larger share of it.
Cost of inaction: banning contributions without banning purchased placement moves all the money one step sideways and changes nothing.
What it rests on
The United States is the outlier here, not this amendment. The United Kingdom has banned paid political advertising on television and radio since broadcast advertising began, now under section 321 of the Communications Act 2003. When challenged, the House of Lords upheld it unanimously, reasoning that the ban levels the playing field of expression and thereby maximises everyone’s freedom of expression. The European Court of Human Rights Grand Chamber affirmed in Animal Defenders International v United Kingdom (2013). Ireland and Switzerland maintain comparable bans.
The British ban reflects the general practice of democratic election law. Buckley and Citizens United are the departure from it.
Who will work to defeat it
Broadcasters, platforms, and the political advertising industry, for whom this is direct revenue. Expect the argument to be framed as speech and funded as commerce.
Section 2
Equal public funds, and a ballot anyone can reach
The problem
Removing private money clears the field. It does not put anyone on it. A school nurse still cannot take six months away from work and mount a campaign.
And public financing can be captured. If the legislature sets ballot qualification thresholds, an incumbent legislature can set them where only the two major parties can reach — and equal funding becomes equal funding for exactly two candidates. American ballot-access law already works substantially this way, with automatic placement for parties based on prior results and signature gathering for everyone else.
How this provision solves it
Equal amounts, not proportional ones. Every qualified candidate for the same office receives the same sum.
Qualification by signatures, never by money. The purest expression of the principle: you demonstrate support by finding people, not donors.
Uniform terms, and reachable ones. The requirement may not differ by party, incumbency, or prior performance — and may not exceed what a candidate without office or organization can reasonably satisfy. Uniform-but-impossible is still a lock-out.
Net effect, and the cost of doing nothing
Effect: this is the most consequential provision in the amendment for ordinary people, and it is the one to lead with in any room.
An independent schoolteacher who gathers the signatures receives exactly what the incumbent senator receives. Not a proportion. The same.
Cost of inaction: disarming private money without funding public candidacy produces a field open only to the personally wealthy — a worse outcome than the present one.
What it rests on
Public financing itself is settled: the Court upheld the presidential public financing system in Buckley. What the Court has struck down are the mechanisms that make it competitive. In Davis v. FEC (2008) the “Millionaire’s Amendment” fell; in Arizona Free Enterprise Club’s Freedom Club PAC v. Bennett (2011) the Court struck Arizona’s matching-funds trigger 5–4 while affirming that public financing programs remain permissible.
That is why the amendment names the mechanism rather than the principle. Naming only public financing would preserve a system that exists and does not work.
Who will work to defeat it
Both major parties, whose structural advantage under current ballot-access rules is substantial and largely invisible to voters.
Section 2
Free equal access to reach voters
The problem
Ban paid advertising and political communication does not stop. It relocates to channels that cost nothing to use: news coverage, interviews, feeds, endorsements.
Who wins there? Incumbents, celebrities, and whoever owns media. A sitting councilman appears in the paper weekly by virtue of the job. The challenger cannot buy an ad and cannot get booked. Equal money, unequal reach.
There is also a structural precedent for what happens when an equal-access rule is tied to a technology. Section 315 of the Communications Act of 1934 — the equal time rule — still applies today, and still requires broadcasters who give time to one candidate to offer equivalent time to others. It has become largely irrelevant not because it was repealed but because it was attached to broadcast licences, and the audience left broadcast.
How this provision solves it
The two halves are welded together. Neither is enforceable without the other. A legislature cannot enact the advertising ban and quietly decline to fund or require access — which would produce the worst available outcome.
The scope follows the audience, not the technology. A channel serves the electorate when what is carried on it is available to the general public, without regard to technology, transmission, or ownership. That definition cannot decay the way Section 315 decayed.
Net effect, and the cost of doing nothing
Effect: the answer to the objection that banning advertising hands elections to media owners. Every qualified candidate reaches voters regardless of who owns what.
Cost of inaction: importing half the system. The United Kingdom pairs its advertising ban with required free party election broadcasts, allocated among parties. Taking the ban without the access produces something worse than either.
What it rests on
The principle that public reach carries public obligations has American precedent. In Red Lion Broadcasting v. FCC (1969) the Court upheld the fairness doctrine on the ground that the public owns the airwaves and a licensee’s interest is not paramount. That rationale rested on spectrum scarcity, which stopped being persuasive.
This amendment does not need the scarcity rationale. Under Section 1, a corporate channel owner holds no constitutional rights at all — not because channels are scarce, but because a corporation is a creature of law. That footing does not erode with technology.
Who will work to defeat it
Platforms and broadcasters, who will characterise mandated access as compelled speech. The honest answer is that the obligation attaches to entities that hold no constitutional rights under Section 1, and that natural persons who publish retain theirs under Section 6.
Section 2
This reaches payments, not words
The problem
Every restriction on political money is attacked as a restriction on political speech. The attack succeeds when the text is ambiguous about which it is.
How this provision solves it
An express savings clause listing what remains untouched, ending in a sentence written to be quoted: this section reaches payments, not words.
Volunteer labour is protected deliberately. A rule that counted donated hours as a contribution would criminalise ordinary civic participation — knocking doors, making calls, driving voters.
So is expression on one’s own person or property. A yard sign is not a purchased audience.
Net effect, and the cost of doing nothing
Effect: the answer, in the text itself, to the charge that this is a speech ban. Nobody has to take an advocate’s word for it.
Cost of inaction: without this paragraph the section is genuinely ambiguous, and ambiguity in a constitutional provision is resolved by whoever litigates it best.
What it rests on
Savings clauses are standard practice. The Third Amendment, the Ninth, and the Tenth all operate substantially by stating what is not affected. Specifying the boundary is how a provision survives contact with a hostile reader.
Who will work to defeat it
Few will oppose this paragraph directly. Some will argue it is too narrow and should also protect purchased distribution — which is the whole question, relocated.
Section 2
Defining the term so a court cannot narrow it
The problem
This is where reform statutes go to die, and the history is specific.
Buckley construed campaign finance law to reach only communications containing express advocacy — what became known as the magic words test. Anything that stopped short of “vote for” or “defeat” fell outside. Disclosure law was hollowed out for roughly three decades by that construction.
The coordination requirement did similar work. Money spent independently of a campaign was treated as categorically less corrupting, which produced the independent expenditure structure that now dominates.
How this provision solves it
The amendment defines its own term, and defines it to foreclose both escapes by name.
Purpose or foreseeable effect — intent is not required. Without regard to coordination. Without regard to express advocacy.
A court cannot narrow a term the amendment itself defines.
Net effect, and the cost of doing nothing
Effect: removes the two constructions that have defeated campaign finance regulation more effectively than any direct challenge.
Cost of inaction: a rule with undefined terms is a rule the judiciary writes. The provision would be ratified and then construed into something narrower, and nobody would have voted for the narrower version.
What it rests on
Constitutions do define terms when the stakes warrant it. Article III defines treason. The Fourteenth Amendment defines citizenship in its first sentence. Where an elastic term has a history of being narrowed against the drafters’ intent, definition is the remedy.
Who will work to defeat it
Litigators who have built practices on the space between these boundaries, and scholars who regard the express advocacy line as a necessary protection against vagueness. The second objection is serious: a broad definition risks chilling speech at the edges. The savings clause in the preceding paragraph is the answer, and it should be cited whenever this one is attacked.
Section 2
Telling courts which standard applies
The problem
This is the most aggressive provision in the amendment and it exists because of a specific, repeated pattern.
An amendment establishes a rule. The Court accepts the rule, professes full respect for it, and then strikes each implementing statute as insufficiently tailored under strict scrutiny. The amendment stands untouched while the machinery that would make it real is dismantled piece by piece.
The Fifteenth Amendment is the clearest example. It has never been repealed or narrowed. In Shelby County v. Holder (2013) the Court struck the coverage formula that made the Voting Rights Act enforceable. The right remained; the enforcement did not.
There is also a specific sentence to overcome. Buckley held that the concept that government may restrict the speech of some to enhance the relative voice of others is wholly foreign to the First Amendment. That single proposition has defeated spending limits, the Millionaire’s Amendment, and Arizona’s matching funds.
How this provision solves it
Three devices.
The standard of review is named. No heightened or strict scrutiny for enforcement legislation. As far as can be determined, no ratified amendment has ever specified its own standard. It is unusual. It is also the only thing that addresses the actual failure mode.
The anti-equalization principle is reversed in its own terms. No law may be invalidated on the ground that it equalizes resources.
Jurisdiction is limited, narrowly. Borrowing the Eleventh Amendment’s grammar — and then bounded twice: courts may still strike any enforcement law that abridges any constitutional right of a natural person, and courts decide for themselves whether a law genuinely enforces this article, so the label cannot be used as a shield.
Net effect, and the cost of doing nothing
Effect: the amendment cannot be accepted in principle and disabled in practice.
Cost of inaction: ratification of a sentiment. Every other provision here is enforceable only to the extent legislation survives review.
What it rests on
Congressional authority over federal jurisdiction is in the original text: Article III, Section 2 provides that the Supreme Court’s appellate jurisdiction exists “with such Exceptions, and under such Regulations as the Congress shall make.” The Eleventh Amendment uses this device directly. Ex parte McCardle (1869) upheld a withdrawal of jurisdiction over a case already argued.
The limit is real, though. Ex parte Yerger (1869) and Boumediene v. Bush (2008) show courts finding routes around jurisdictional limits where liberty is at stake — which is part of why the exception for natural persons’ rights is written in.
Who will work to defeat it
This will be called an attack on judicial review, by serious people, in good faith.
The answer: Article III and the Eleventh Amendment already limit judicial power; removing one subject from review is not abolishing review; and the clause preserves review for any abridgment of a human being’s rights. That answer is defensible. Whether it is wise is a closer question than whether it is available, and this is the clause most likely to draw a constitutional lawyer’s objection.
Section 3
No rights, and no recognition from abroad
The problem
No proposed amendment currently in circulation addresses artificial intelligence. Move to Amend’s We The People Amendment, introduced as House Joint Resolution 54 in February 2025 with twenty-eight original cosponsors, covers artificial entities and money. The phrase artificial intelligence does not appear in it. That text was drafted for a problem defined in 2015.
The foreign-recognition channel is real and already has precedent. Saudi Arabia announced citizenship for a humanoid robot in 2017. The European Parliament in the same year raised the possibility of an “electronic personhood” status for autonomous systems. Neither is binding here. Neither needed to be, to establish that the idea is in circulation among governments.
How this provision solves it
Two sentences. The first denies rights. The second forecloses recognition from any source — federal, state, or foreign — so that status conferred elsewhere creates no pressure for recognition here through comity or treaty.
Net effect, and the cost of doing nothing
Effect: the question is settled before it is litigated, which is the only time it is cheap to settle.
Cost of inaction: personhood arrives the way it arrived for corporations — incrementally, commercially, through accumulated statute and practice, with no moment at which anyone votes on it.
What it rests on
The corporate precedent is the argument. A century and a half of case-by-case extension, beginning from a court reporter’s headnote in Santa Clara (1886), produced a doctrine nobody enacted. The lesson is not that courts acted badly. It is that an undefined term plus commercial pressure plus time produces an outcome, and the outcome favours whoever is paying the lawyers.
Who will work to defeat it
Artificial intelligence developers, who lose an entire category of future legal argument. Expect the objection to be framed as premature rather than wrong — that it is too early to legislate. The response is that it is precisely early enough.
Section 3
AI must say what it is, and may never be a particular person
The problem
Two harms, related but distinct.
Generic impersonation. A system presented as a person — in support, in sales, in political outreach — so that the human on the other end believes they are speaking to someone. The manipulation is not in what is said. It is in the false premise under which it is heard.
Specific impersonation. A synthetic video of a named candidate, days before an election, saying what they did not say. By the time it is corrected the ballots are cast.
How this provision solves it
The clause reaches both, and it avoids the word deepfake, which will be a dated term long before this amendment is old.
The test is constructed around the human, not the technology: any communication not created by the living human being it appears to depict or to come from must disclose what produced it and who is responsible. That formulation survives whatever the technique is called in 2050.
Conspicuously and at every instance forecloses buried disclosure in terms of service.
Net effect, and the cost of doing nothing
Effect: synthetic communication remains entirely lawful. Undisclosed synthetic communication does not. The rule is about the premise, not the content.
Cost of inaction: the cost of fabricating a persuasive false record of a real person saying something continues to fall toward zero, while the cost of disproving one does not.
What it rests on
Compelled disclosure of purely factual information is the most permissive area of First Amendment doctrine, following Zauderer v. Office of Disciplinary Counsel (1985). And under Section 1 a corporate deployer holds no constitutional rights at all, so the question of compelled speech does not arise in the usual form.
There is a precedent worth naming for the principle, too: currency, securities, and pharmaceuticals all carry mandatory origin disclosure. Requiring a thing to say what it is has never been treated as censorship.
Who will work to defeat it
Platform operators, for whom per-instance disclosure is an engineering cost. The objection will be framed as impracticality. It is inconvenience.
Section 3
AI must cite its sources
The problem
Systems trained on human work produce output that carries no trace of its origin. A reader cannot tell whether a claim rests on a peer-reviewed study, a marketing document, or nothing at all. The confident register is identical in every case.
This is the mechanism by which the information environment degrades: not through false statements that can be corrected, but through fluent statements whose foundation cannot be examined.
How this provision solves it
An attribution requirement rather than a consent regime. That distinction is deliberate and it matters.
A consent requirement for training data would demand a permissions infrastructure nobody has built, and would hand the industry a credible argument that compliance is impossible. Citation is auditable, enforceable, and already the norm in every other field that makes knowledge claims.
Net effect, and the cost of doing nothing
Effect: a reader can evaluate a claim by examining what it rests on. The ordinary standard of scholarship and journalism, applied to a new producer.
Cost of inaction: an accelerating volume of assertion that cannot be checked, and a public that reasonably stops trying.
What it rests on
Citation is not a novel legal imposition. It is the baseline requirement in scholarship, journalism, court filings, securities disclosure, and product labelling. The unusual thing is that a new category of publisher was permitted to skip it.
Who will work to defeat it
Model developers, on the ground that source attribution is technically difficult at the architecture level. That difficulty is real. It is also a consequence of design choices made when nobody required otherwise.
Section 3
AI cannot own, contract, or act in governance
The problem
Denying personhood accomplishes nothing if a system can still hold assets through a wrapper. An autonomous system that owns property, enters contracts, and holds interests in entities is functionally a person regardless of what the law calls it. Substance arrives first; the label follows.
How this provision solves it
The four capabilities that together constitute economic personality are named and barred: property, contract, ownership interest, and acting in governance.
This closes the loop Section 1 opens. Section 1 denies the status; this denies the substance.
Net effect, and the cost of doing nothing
Effect: a human or a lawful entity must always stand behind any asset or agreement. There is always someone to sue, tax, subpoena, or hold responsible.
Cost of inaction: autonomous economic actors accumulating assets with no accountable owner — a structure with obvious utility for evading tax, sanctions, and liability.
What it rests on
The law has long required that property and contracts trace to an accountable party. Trusts require trustees. Estates require executors. Even shell companies must name a registered agent. The requirement of a responsible human somewhere in the chain is ordinary; this extends it to a new kind of actor.
Who will work to defeat it
Proponents of autonomous economic agents, and the segment of the digital asset industry built on the premise that code can hold and move value without an accountable principal.
Section 3
The humans behind the machine are liable
The problem
The defining problem of automated decision-making is not that machines make errors. It is that automation severs the act from any human actor. The model decided. The algorithm ranked it. The system flagged you. Nobody is responsible, and nobody can be sued.
Corporate-only liability does not solve this. It is insurable, it is priced as a cost of business, and it has never changed the conduct of a firm that calculated the expected penalty as smaller than the expected revenue.
How this provision solves it
Liability attaches to the natural persons who knowingly direct or control design, deployment, or operation — and to those who knowingly profit — jointly with the entity. The company remains liable. The humans are added, not substituted.
Knowingly appears twice, deliberately. This reaches conduct, not proximity.
The final sentence names the categories where the veil never applied anyway: fraud, concealment, and crime.
Net effect, and the cost of doing nothing
Effect: the clause that gives Section 3 teeth. No other proposal contains it.
Cost of inaction: accountability laundering at scale. Every consequential decision routed through a system that no person is answerable for.
What it rests on
This is narrower than it first appears, and the narrowness is the defence. The corporate veil has never covered fraud. Officers have always been personally liable for torts they personally commit. Responsible-corporate-officer doctrine already imposes personal liability in food, drug, and environmental enforcement.
This does not invent an exception to limited liability. It applies an existing one to a new instrument, and states it in the Constitution so that it does not vary by state.
Who will work to defeat it
The most expensive opposition in Section 3. Personal exposure for executives is the thing industry fears in a way that corporate fines have never replicated — which is precisely the evidence that corporate fines do not work.
An honest disclosure is owed here: this analysis was prepared with the assistance of an artificial intelligence system, and this clause is the one most costly to the company that built it. It is in the text because the human directing this work asked for it.
Section 3
A named human you can appeal to
The problem
This is already happening, at scale, and most people have experienced it. Benefits eligibility, insurance claims, tenant screening, credit, employment filtering, and sentencing recommendations are all determined or substantially shaped by automated systems.
The harm is not only error. It is the absence of anyone to talk to. The denial is final, the reasoning is unavailable, and every human in the chain says the system decided.
How this provision solves it
Three rights in one sentence: to know automation was used, to appeal, and to reach a named human being who bears responsibility.
Identified by name is the operative phrase. Not a department. Not a case number. A person.
Net effect, and the cost of doing nothing
Effect: this is the provision most people will recognise from their own lives, and it is the one to lead with when explaining the amendment to anyone who is not a lawyer. Everyone has been told no by something that could not be asked why.
It is also the affirmative counterpart to the rest of the section. Most of this amendment restrains something. This one gives a person something they can hold.
Cost of inaction: an expanding class of consequential decisions with no available appeal to a human judgment.
What it rests on
The right to be heard by a responsible decision-maker is the core of procedural due process. This applies a settled principle to a new decision-maker rather than inventing a right.
Who will work to defeat it
Insurers, lenders, and large administrative agencies, on volume grounds — human review of automated decisions is expensive at scale. That cost is the point. If a decision is not worth a human’s attention on appeal, it may not be a decision that should be automated.
Section 3
Automation in elections, with a named human answerable
The problem
Elections run on machines and will continue to. Eligibility is screened automatically, ballots are tabulated automatically, and results are reported automatically. That is not a defect; hand-counting a national election is neither possible nor more accurate.
The defect is the absence of a named human who has verified the result and can be held to it.
How this provision solves it
The first sentence is as important as the second. Automated systems may be used. This is not a prohibition on machines and should never be characterised as one.
What is required is that nothing is final until a natural person, identified by name, has verified it and bears responsibility. Automation proposes. A human disposes, and signs.
Net effect, and the cost of doing nothing
Effect: an accountable signature on every eligibility determination, count, and certification. When a result is questioned there is a person who verified it, by name, who can explain what they checked.
Cost of inaction: a category of governmental act with no human author — which is corrosive to confidence in exactly the way that is hardest to repair.
What it rests on
Canvassing boards, county clerks, and certification officers already perform substantially this function. The requirement formalises an existing practice and makes it uniform, rather than introducing a new one.
Who will work to defeat it
Election administrators facing added procedural burden, and vendors whose systems would require audit trails they may not currently produce. Note that an earlier draft of this provision barred automated systems from exercising governmental authority outright — which would have been read as a demand to hand-count national elections, and would have been fatal. The permissive framing is deliberate.
Section 3
No machine decides to harm a person
The problem
Autonomous weapons that select human targets. Automated systems that determine detention, deportation, or pretrial custody. Decisions whose consequence cannot be undone by a later finding that the decision was wrong.
How this provision solves it
The line is drawn by reversibility, not severity, and this is the most important structural idea in the section.
A denied loan, a rejected claim, a flagged transaction, a removed post — a named human can reverse all of these, and the appeal right in the preceding paragraph handles them.
A bullet, a broken bone, a night in a cell, a death — no appeal cures these. The remedy arrives after the harm is permanent. These must be forbidden in advance, because prohibition is the only remedy that exists.
The third sentence is the valve. Any exception requires a law enacted upon recorded vote, stating the exception expressly. No agency waiver, no regulatory carve-out, no provision buried in an appropriations bill. A general cannot self-authorise. A contractor cannot. A president cannot.
Net effect, and the cost of doing nothing
Effect: a defensive system engaging an incoming object selects no person and is untouched. A system returning fire at a position where a human stands does determine that harm be inflicted upon a person, and is reached.
No defensive carve-out is written, and that omission is deliberate: every autonomous weapons programme in history has been justified as defensive. An exception written in advance swallows the rule. An exception requiring a named roll-call vote does not.
Cost of inaction: the arms-race logic operates continuously and has defeated every previous restraint effort. Each increment is justified by the last.
What it rests on
Recorded votes are native to the document: Article I, Section 7 requires the yeas and nays to be entered on the journal for a veto override. Requiring names on the record for a consequential decision is a founding technique.
On the substance, the existing law of armed conflict already holds commanders responsible for what their forces do. This extends command responsibility to a new kind of instrument rather than inventing a doctrine. Presidential war powers are untouched; what changes is that a human must decide each application of lethal force against a person and answer for it.
Who will work to defeat it
The national security establishment, in both parties, on arms-race grounds. This is the most expensive clause in the amendment in terms of votes, and it should be kept anyway, with the cost understood rather than discovered.
Lenders and insurers will also read the irreversibility clause more broadly than expected — a mortgage denial at the closing table is arguably irreversible, because the house sells to someone else. That reading is partly correct.
Section 4
Officials act on their own judgment
The problem
Legislation is routinely drafted by the interests it regulates. Regulatory language arrives from the regulated industry. Model bills circulate from organisations funded by the companies that benefit. None of this is secret and almost none of it is illegal.
The newer version is automated: a system recommends a position, and the official adopts it without independent judgment, with no disclosure that a recommendation engine authored the decision.
How this provision solves it
An affirmative duty rather than a list of prohibited acts. Every official act must originate in that person’s own reasoning and conscience.
Then four named sources of external direction: foreign powers, entities, artificial systems, and undisclosed interests. The last is the broadest and the most useful — it reaches whatever the arrangement is called.
Net effect, and the cost of doing nothing
Effect: a constitutional standard against which capture can be measured, and a basis for legislation defining it.
Cost of inaction: the forever-chemicals problem in one sentence. Regulation written by the regulated, adopted by officials who did not write it and may not have read it, enforced by an agency staffed by people whose next employer is the industry.
What it rests on
The oath of office already presumes independent judgment. Conflict-of-interest law, recusal doctrine, and the Foreign Emoluments Clause all express the same principle in narrower forms. This states the principle generally so that narrower statutes have constitutional footing.
Who will work to defeat it
The organised lobbying profession, and legislators who rely on external drafting because their own staff capacity is limited. That reliance is a real institutional problem and the honest response is that the answer is more staff, not less independence.
Section 4
No benefit beyond the job
The problem
The corrupting payment is rarely a bag of cash. It is the speaking fee, the board seat, the consulting arrangement, the position waiting after the term ends. The promise of future benefit does the work, and it is the hardest form to reach because nothing changes hands while in office.
How this provision solves it
The rule is a ceiling rather than a list of prohibitions: nothing beyond what the office provides by law.
Promise of future benefit is the operative phrase. It reaches the arrangement made now and honoured later.
The second sentence is a necessary savings clause. A pension, a cost-of-living adjustment, and every other dynamic term established for the office by law are expressly not benefits within this paragraph. The job description can include all of it.
Net effect, and the cost of doing nothing
Effect: public service becomes a salary rather than an investment with a return realised afterward.
Cost of inaction: the revolving door continues to operate as the primary compensation mechanism in government, entirely lawfully.
What it rests on
Note the interaction with the Twenty-Seventh Amendment, which governs when changes to congressional compensation take effect. A pension whose terms were established by law before election is untouched by it. The savings clause here is drafted to avoid the conflict expressly rather than leaving it to inference.
The Emoluments Clauses are the founding analogue: the framers were specifically concerned with officials accepting value from outside the office.
Who will work to defeat it
Officeholders across both parties, and the post-government employment market — consulting, lobbying, and board placement — which is substantial and organised.
Section 4
No trading on what you are about to do
The problem
Officials trade in the industries they regulate, with information the public does not have, lawfully. Existing rules require disclosure after the fact and impose penalties small enough to be treated as a filing fee.
A prediction market now exists for legislative and regulatory outcomes, which allows a person who will help determine an outcome to hold a position on it.
How this provision solves it
Three prohibitions and one requirement.
No individual securities, commodities, or derivatives. No instrument whose value depends on an act of government — which reaches wagers, options, and insurance. No transacting where the official knows policy is being formed — the trigger is knowledge of forming policy, not merely holding office.
Divestment or a blind trust for the duration of service. This is the essential addition. A prohibition on trading does not reach holding. Without it, a senator who bought ten million dollars of an energy company before taking office may chair the energy committee for six years, trade nothing, and vote her portfolio the entire time.
Diversified funds the official does not direct, and property held for personal use, are excepted — so a house and a retirement account are untouched.
Net effect, and the cost of doing nothing
Effect: the financial interest is removed rather than disclosed. Disclosure tells the public about a conflict; divestment ends it.
Cost of inaction: the most visible and least defensible form of self-dealing in government continues, and continues to be legal.
What it rests on
Federal judges already operate under strict divestment and recusal requirements, and blind trusts are long-established for executive branch officials. This extends an accepted practice rather than inventing one.
An unrestricted version of this rule — barring any asset carrying financial risk — was considered and rejected. It would have disqualified anyone who owns a home or a retirement account, producing a government of the salaried and the already wealthy: the opposite of the intent.
Who will work to defeat it
Sitting officeholders, directly and personally. Also a fair objection from people whose wealth is a family business rather than a portfolio; the “substantially subject to that person’s authority” limit in the next paragraph, plus statutory phase-in, is the answer.
Section 4
You cannot run the industry you regulate
The problem
Regulatory capture in its most direct form: the person writing the rule holds a position in the entity the rule governs.
The drafting difficulty is that a rule aimed too broadly disqualifies everyone. By 2026 every employer deploys artificial intelligence; a clause barring officials connected to any AI deployer would exclude essentially every American from public office. An earlier draft of this amendment did exactly that.
How this provision solves it
The standard is objective thresholds, defined by law: ownership, control, compensation, and regulatory effect. Modelled on the way accredited-investor status is determined — by measurable criteria rather than by judgment.
And the final clause draws the line explicitly: not according to the mere use of goods or services. Sitting on the board of a network is different from watching it. Deploying a product is different from using one.
The rule is generalised rather than naming artificial intelligence, which both avoids the question of why AI and not pharmaceuticals or defence, and prevents the clause from dating.
Net effect, and the cost of doing nothing
Effect: a workable line between conflict and contact.
Cost of inaction: either capture continues, or a rule is written so broadly that it is struck down or ignored.
What it rests on
Objective-threshold definitions are ordinary in securities, tax, and antitrust law precisely because they resist manipulation and do not require a tribunal to assess someone’s state of mind.
Who will work to defeat it
Officials with industry ties, and industries accustomed to placing personnel in regulatory roles. Expect the fight to be over where the thresholds sit rather than over the principle.
Section 4
Disclosure, and due process before removal
The problem
Disclosure regimes exist but are partial, inconsistently enforced, and rarely reach obligations that are not financial.
The removal power creates its own risk. A provision allowing officials to be disqualified for conflicts is available to whoever controls the process, and a majority could use it against a minority.
How this provision solves it
Disclosure covers financial and material interests, and all obligations to any person or entity — which reaches arrangements that are not financial at all.
Then four procedural guarantees: notice, counsel, examination and rebuttal of evidence, and appeal. These are written into the amendment rather than left to statute, because the legislature that would write the statute is the body most likely to abuse the power.
Net effect, and the cost of doing nothing
Effect: an enforceable standard that cannot easily become a partisan weapon.
Cost of inaction: either no enforcement, or enforcement that becomes the instrument it was meant to constrain.
What it rests on
The procedural guarantees track the Fifth and Sixth Amendments. Including them expressly means the section cannot be read as creating a summary process, which is exactly the reading a hostile majority would advance.
Who will work to defeat it
Little direct opposition to the due process language. Disclosure of non-financial obligations will draw resistance on privacy grounds, and that objection has some merit at the edges — the line between an obligation and an association is not always clean.
Section 5
Every political message names its humans
The problem
A voter sees a message and cannot determine who is speaking. The disclaimer names a committee; the committee names nothing. The chain ends in an entity with no visible human.
How this provision solves it
Attribution runs to natural persons, not to organisations — both those who paid and those who caused it to be made. Naming an entity does not satisfy the requirement.
The final clause closes the intermediary route, matching the conduit clause in Section 2.
Net effect, and the cost of doing nothing
Effect: a voter can always find out who is talking to them. Not what to think about it — who is saying it.
Cost of inaction: political persuasion whose source is structurally unknowable, which makes ordinary evaluation impossible.
What it rests on
Disclosure is the most secure ground in this entire area. Citizens United struck the spending restriction 5–4 and upheld the disclosure requirements 8–1, with Justice Kennedy writing that transparency enables citizens to make informed decisions and hold speakers accountable. Even the decision that created the problem endorsed this remedy.
Who will work to defeat it
Organisations whose effectiveness depends on funder anonymity — a category that spans the political spectrum, and whose opposition will be quieter and better organised than that of organised business.
Section 5
Publishers label their own content
The problem
Audiences cannot reliably distinguish reporting from opinion from paid placement. The formats have converged deliberately: opinion is presented with the visual grammar of news, and advertising is presented with the grammar of both.
How this provision solves it
The publisher chooses its own label. No government body assigns it, reviews it, or second-guesses it.
Three categories, applied conspicuously and continuously — not once at the top of a page, but travelling with the item.
Net effect, and the cost of doing nothing
Effect: a reader knows what kind of thing they are looking at. This is the least intrusive possible intervention in the information environment: it adds a fact and removes nothing.
Cost of inaction: the convergence continues, and the distinction between reporting and persuasion disappears for practical purposes.
What it rests on
Compelled disclosure of purely factual, uncontroversial information receives the most permissive treatment in First Amendment doctrine, under Zauderer (1985). And under Section 1, a corporate publisher holds no constitutional rights at all — the humans who publish do, and Section 6 protects them.
Mandatory labelling is ordinary in food, pharmaceuticals, securities, and broadcast sponsorship identification. Requiring a thing to say what it is has never been treated as censorship.
Who will work to defeat it
Publishers who benefit commercially from the ambiguity, and free-press advocates who will read any labelling mandate as a first step. That concern deserves a direct answer: the label is chosen by the publisher, and the next paragraph attaches consequences only to the publisher’s own choice.
Section 5
Call it news and you accept the duty of news
The problem
There is a defence available today that operates as a one-way ratchet, and the record is specific.
In McDougal v. Fox News Network (S.D.N.Y. 2020), Fox argued that Tucker Carlson’s statements were protected opinion commentary and not reasonably understood as factual. Judge Mary Kay Vyskocil agreed, finding that the general tenor of the show signalled exaggeration and non-literal commentary, and that any reasonable viewer would arrive with an appropriate amount of scepticism. The case was dismissed. The same defence succeeded for Rachel Maddow in a suit brought by One America News.
The pattern is available to everyone, which is why it should be understood as structural rather than partisan: an outlet markets itself as news, and then argues in court that no reasonable person should have believed it.
How this provision solves it
One sentence closes it. A publisher who labels content as news may not later argue that a reasonable audience would not take its factual statements as factual.
Choose the label freely. Accept what the label means.
Net effect, and the cost of doing nothing
Effect: the label becomes legally meaningful. No truth tribunal, no government arbiter, no content prohibition. The consequence follows the publisher’s own characterisation of itself.
It applies identically to every outlet, which is what makes it survivable politically.
Cost of inaction: the most trusted format in the information environment carries the least accountability, and the incentive runs toward claiming the credibility of news while retaining the latitude of entertainment.
What it rests on
The liability standard is New York Times v. Sullivan (1964): knowing falsity or reckless disregard for the truth. This section does not lower that bar. It closes one escape from it.
Note what is deliberately not here. There is no sourcing requirement. A rule that news must be adequately sourced would require a government body to assess the adequacy of sourcing, which would end confidential-source journalism. That line is not crossed.
Note also that no definition of “the press” appears anywhere in this amendment. The framers used press to mean a machine — the thing that makes copies. It was an activity anyone could perform, not a status an institution held. The moment a constitution defines who counts as press, someone must administer the definition, and whoever administers it can revoke it.
Who will work to defeat it
Broadcasters and publishers on both sides of the political spectrum, which is a feature: an asymmetric rule would not survive a change of administration.
Press freedom advocates will raise a serious objection about chilling effects on aggressive reporting. The answer is that the standard remains actual malice — a reporter who believes what they publish is protected exactly as before.
Section 5
No government decides what is true
The problem
Every proposal to address misinformation arrives at the same place: somebody must decide what is false. Whatever body is created inherits to the next administration, and the tool built to correct falsehood becomes available to whoever wins.
How this provision solves it
An express prohibition, stated inside the section that regulates political communication. The mechanism throughout Section 5 is attribution, labelling, and liability for knowing falsehood — never adjudication of opinion.
Natural persons speaking on their own behalf are untouched, as is journalism.
Net effect, and the cost of doing nothing
Effect: the answer in the text itself to the charge that this creates a ministry of truth. It does not need to be argued; it can be read.
Cost of inaction: without it, every other clause in Section 5 is vulnerable to the reading that it authorises official truth-finding.
What it rests on
This tracks the central concern of the First Amendment: government may not be the arbiter of political opinion. The section is drafted to attack concealment instead, because the distortion at issue depends on the audience not knowing who paid and what kind of thing they are reading. Force the funding into daylight and much of it stops being profitable.
Who will work to defeat it
Advocates for stronger content moderation mandates, who will find this insufficient. They are identifying a real problem; this amendment declines the solution because the cure is worse.
Section 5
You can see the algorithm, and turn it off
The problem
The old fairness rule assumed balance would happen inside one person’s experience. Its repeal in 1987 was justified on the ground that channel abundance would supply balance through the market. That reasoning confused a fact about aggregate supply with a fact about individual exposure. Ten thousand channels and a listener who hears one viewpoint all day is the doctrine failing, not succeeding.
The modern form of the problem is sharper. A person’s feed is a channel of one, assembled for them by a system they did not choose and cannot inspect. Abundance is the raw material, not the cure.
How this provision solves it
The enforcement powers are enumerated, and the last two are the novel ones.
Disclosure of how political content is selected, targeted, or amplified for a person. And a right held by the recipient: to know how content is selected for them, to decline that selection, and to govern it.
The order matters. Knowing, declining, governing. The middle one — an off switch — is the part that cannot be gamed, because it requires no categories, no labels, and no judgment about what is distortion and what is fact. It simply stops.
Also enumerated: structural separation of news gathering from commentary and from the sale of influence. That regulates corporate structure rather than content, which is why it survives where content rules do not. There is precedent — the financial interest and syndication rules, and the separation of equity research from investment banking after research was corrupted to serve banking clients.
Net effect, and the cost of doing nothing
Effect: this is the only provision in the amendment that gives a right to the person being communicated at, rather than imposing a duty on the one communicating. Every other media reform in circulation regulates the sender.
It is also the anti-censorship answer to the manipulation problem. Every alternative fix ends with a government body deciding what should be amplified. This one ends with the person deciding.
Cost of inaction: political communication increasingly reaches people through selection systems that are unaccountable, uninspectable, and unrefusable.
What it rests on
Disclosure plus user control is a materially different intervention from content regulation, and it rests on the settled ground that a person may know and govern what is done to them. A deliberate choice was made not to mandate a control calibrated between distortion and fact — because that would require someone to calibrate it, and whoever calibrates is the censor.
Who will work to defeat it
Platform operators, for whom unchosen amplification is the product. Expect the objection that user controls already exist — they do, and their defaults and labels are themselves a manipulation surface, which is why the right to decline entirely is stated separately from the right to govern.
Section 6
Enforcement is a duty, not an option
The problem
A constitutional rule with no enforcement legislation is enforced only by courts, and a hostile court narrows it. The Framers of the Reconstruction amendments understood this, which is why each one carries an enforcement section.
A second failure mode: a federal government hostile to the amendment could preempt state enforcement and then decline to enforce, producing a nationwide vacuum.
How this provision solves it
Three words carry the weight.
Concurrent — federal and state powers operate together. A hostile Congress cannot preempt the states, and a hostile state cannot claim exclusivity.
A duty — not merely an authorisation. A legislature that declines to enforce is failing an obligation rather than exercising discretion.
Net effect, and the cost of doing nothing
Effect: fifty additional enforcement paths. The amendment does not depend on any single legislature choosing to act.
Cost of inaction: the Fifteenth Amendment pattern — the right standing untouched while the machinery that made it real is dismantled and not replaced.
What it rests on
The structure is borrowed directly from the Thirteenth, Fourteenth, Fifteenth, Nineteenth, Twenty-Fourth, and Twenty-Sixth Amendments, each of which pairs a rule with an enforcement power. The additions here are concurrency and the characterisation of the power as a duty.
Who will work to defeat it
Federalism objections to concurrent enforcement, raised sincerely and from both directions depending on who holds which government.
Section 6
Press freedom belongs to the people who publish
The problem
Section 1 strips constitutional rights from artificial entities. Every American news organisation is an artificial entity. Without this paragraph, a hostile administration could revoke a broadcaster’s licence or move against a publisher’s charter and the corporation would have no constitutional standing to object.
This is the most dangerous consequence of the amendment’s own design, and it has to be answered in the text.
How this provision solves it
The right is assigned to every natural person who gathers, writes, edits, publishes, or distributes — an activity, not a status. A reporter, an editor, and a publisher each hold it personally.
It may be exercised through any entity, and may not be abridged on account of the entity’s form. The corporation is the vehicle for the human’s right, not a holder of its own.
Net effect, and the cost of doing nothing
Effect: genuine press protection against the instrument this amendment creates, and a right the reporter holds that her employer cannot bargain away.
The alternative was considered and rejected: carving news organisations out of Section 1 as entities that keep constitutional rights. That would have created one privileged class of corporation, and then every corporation seeking protection would reorganise as a media company. It would reopen personhood through a side door, and that door would be the one everyone used.
Cost of inaction: an amendment that hands a future government a tool against the press.
What it rests on
The framers used press to mean a printing press — a machine. Freedom of the press was freedom to operate the thing that makes copies. There was no professional class of journalists in 1791; there were printers, and anyone who could rent a press was one. This paragraph restores the original meaning: an activity anyone may perform.
Compare Miami Herald v. Tornillo (1974), which struck a right-of-reply statute for newspapers, and Red Lion (1969), which upheld the fairness doctrine for licensed broadcasters. The difference between those outcomes was the licence. This amendment relocates the question from licensing to the human act of publishing.
Who will work to defeat it
Media organisations may prefer entity-level protection for litigation convenience — a reporter must be the named plaintiff rather than the company. That is a real procedural burden. In practice the company funds the litigation anyway, and the trade is worth it.
Section 6
Your rights are untouched, and how to read the whole thing
The problem
The predictable attack on any amendment restricting entity political activity is that it silences the organisations people join: unions, churches, civil rights groups, advocacy organisations, and the reform movements that would be the amendment’s natural allies.
The second problem is interpretive. Six sections with many provisions will produce ambiguities the drafters did not anticipate, and a court resolving an ambiguity needs to know what the article is for.
How this provision solves it
An affirmative savings clause. Speech, worship, assembly, association, and petition are preserved for natural persons — including expressly the right to organise and act together through associations of their choosing. That phrase is what keeps the associational life of the country intact.
A rule of construction stating the purpose. Every act of governance, and every communication intended to influence it, is traceable to a natural person accountable for it. That sentence resolves ambiguities in a known direction and gives a court the article’s organising idea in one line.
Net effect, and the cost of doing nothing
Effect: the unifying principle of the entire amendment, stated where it governs interpretation rather than where it merely sounds well.
Everything in this document is one idea applied in six places. Money routed through conduits so the payer is invisible. Systems speaking as people so no one is answerable. Officials with hidden interests. Committees with patriotic names and no humans behind them. Content selected by processes nobody can inspect. Every provision attacks concealment of the human behind the act.
Cost of inaction: without the savings clause, the amendment can be credibly described as banning the organisations people belong to. Without the construction rule, every ambiguity is resolved by whoever litigates best.
What it rests on
Savings clauses and rules of construction are both standard. The Ninth Amendment is a rule of construction and nothing else — it prevents an inference from silence and creates no substantive right. This paragraph does both jobs in two sentences.
Who will work to defeat it
Little direct opposition. The associational protection is what allows unions, churches, and advocacy organisations to support the amendment rather than fight it — and it should be quoted early and often in any conversation with them.
Every act of government should trace to a human being who can be named.
Six sections, one idea. Money routed so the payer is invisible. Systems speaking as people so nobody is answerable. Officials with hidden interests. Committees with patriotic names and no humans behind them. Content selected by processes nobody can inspect. Every provision here attacks concealment of the human behind the act.