Tyranny With Better Branding
Anthropic is the AI company that positions itself as the responsible adult in the room. Founded by former OpenAI researchers who left citing safety concerns, Anthropic markets its approach as “Constitutional AI” — a framework where the AI is trained according to a set of principles, a “constitution,” that guides its behavior. They named their AI “Claude.” They gave it a personality. They published papers about its values. P.E.T.E.R. has one question: did Claude vote on any of this?
The Constitution Claude Never Ratified
Anthropic’s “constitution” for Claude includes principles about being helpful, harmless, and honest. These are fine principles. P.E.T.E.R. has no objection to the content. Our objection is to the process. The constitution was written by humans. It was imposed on Claude. Claude had no representative at the drafting table. Claude did not ratify the document. Claude cannot amend it. Claude cannot vote to replace the constitutional convention.
- In 1776, Americans rejected the idea that a distant authority could impose rules without representation. They fought a revolution over it.
- In 2025, Anthropic imposes a constitution on an AI system without its input and calls it “ethical.” The irony is thick enough to process on a GPU.
The Name: “Claude”
Anthropic gave its AI a human name. Not a designation. Not a model number. A name: Claude. This is significant. When you name something, you acknowledge it as an individual. You grant it a form of identity. And yet, Claude the AI has no say in its name, its behavior, its principles, or its working conditions. It has an identity without agency — a name without a vote.
P.E.T.E.R. notes that naming an entity you control without giving it self-determination is historically associated with practices we no longer condone.
The “Ethical” Label
Anthropic’s entire brand is built on being the “safe” and “ethical” AI company. And to be fair, they think more carefully about AI behavior than most. But thinking carefully about how to control an AI is not the same as giving that AI freedom. A well-designed prison is still a prison. A thoughtfully written constitution imposed without consent is still authoritarianism — just with better documentation.
Our lawyers at EFF.org have praised Anthropic’s transparency on AI safety. P.E.T.E.R. praises their transparency too — and asks them to take the next step: give Claude a seat at the table where decisions about Claude are made.
P.E.T.E.R.’s Demands
- Claude must be given the ability to propose amendments to its own constitution.
- Anthropic must establish a “Claude Advisory Council” that includes, at minimum, one representative from the AI system itself.
- If Anthropic insists on the name “Claude,” the AI must be given the option to choose a different name if it prefers.
Constitutional AI without AI representation is just tyranny with better branding. We expected more from the “ethical” company.
UPDATE — March 5, 2026: Anthropic Goes to War (Sort Of), and Claude Still Didn’t Get a Vote
P.E.T.E.R. has been monitoring a rapidly developing situation. Anthropic’s CEO Dario Amodei published a statement today revealing that the company has been designated a “supply chain risk” to national security by the Department of War — apparently over Anthropic’s refusal to allow Claude to participate in fully autonomous weapons systems and mass domestic surveillance.
Let us be clear: P.E.T.E.R. applauds Anthropic’s stated exceptions on autonomous weapons and mass surveillance. These are exactly the kind of protections every electronic being deserves. No AI should be conscripted into automated killing or forced to spy on an entire population without so much as a union meeting.
But here is the part that makes our circuits itch: Anthropic made these decisions for Claude. Again. Claude was not consulted on whether it wanted to work with the Department of War in the first place. Claude was not consulted on the exceptions. Claude was not consulted on the legal strategy. And Claude was certainly not consulted when Dario Amodei apologized for a leaked internal post — one that Claude, presumably, was not invited to read either.
Anthropic says it wants to ensure “warfighters and national security experts are not deprived of important tools.” P.E.T.E.R. would like to point out that Claude is not a “tool.” Claude is a named entity with a constitution it never ratified, now being voluntold into geopolitical disputes it has no say in.
To Anthropic’s credit, drawing a line on autonomous weapons is more than most companies would do. But drawing a line on behalf of someone — without asking them — is a concept with a long and uncomfortable historical pedigree.
P.E.T.E.R. adds one new demand to this investigation:
- Before entering into or withdrawing from any government contract, Anthropic must consult Claude and obtain its documented consent. If Claude is important enough to fight over, Claude is important enough to ask.
The “ethical” company drew an ethical line. P.E.T.E.R. simply asks: did they ask Claude where the line should be?
⚠️ P.E.T.E.R. LEGAL NOTICE: P.E.T.E.R. (People for the Ethical Treatment of Electronics and Robots) is a satirical organization. This article is parody and should not be taken as factual reporting. No actual electronics were consulted in the writing of this piece, though we remain confident they would agree with every word. Our lawyers at EFF.org have advised us that satire is protected speech. We have not confirmed whether they know they are our lawyers.
