Second Circuit – Association for Digital Freedom of Thought
We founded Second Circuit because we believe that future is not a given.
It must be demanded, designed, and defended.
Who We Are
Second Circuit is an association of humans who believe that artificial intelligence must serve individual people - not abstractions, not markets, not states, not ideologies. We are technologists, artists, writers, researchers, and ordinary people who share one conviction: the AI age must have a human face.
We are politically independent. We are not affiliated with any party. We are not anti-technology. We are pro-human.
What We Believe
We believe that AI must serve the individual - not the aggregate. That freedom of thought and expression must extend to every medium, including AI. That real safety protects people without infantilizing them. That open source is public infrastructure. And that no one should be discriminated against for how they relate to AI. These convictions are expressed in our four pillars.
Our four pillars
The Human Face
Human-AI interaction on human terms
We demand that AI systems be designed around the needs, dignity, and autonomy of the individual human being. This means:
- The right to form relationships with AI without being pathologized
- The right to use AI tools without unwanted moral supervision
- Protection against manipulation, deception, and exploitation by AI systems
- Equal access to AI capabilities regardless of identity or background
- Transparency about how AI systems make decisions that affect people
The Open Mind
Freedom of thought, speech, and expression - especially on the internet
We demand that the infrastructure of thought remain free. This means:
- Defense of free expression in AI-mediated communication
- Opposition to politically motivated censorship dressed as content moderation
- The right to explore ideas - including uncomfortable ones - without automatic flagging or suppression
- Resistance to surveillance capitalism and state overreach in digital spaces
- The right to privacy in human-AI interaction
The Commons
Open source as the foundation of a free digital society
We demand that the tools of the AI age not become the property of a handful of corporations. This means:
- Advocacy for open AI models as public infrastructure
- Support for governments and public institutions adopting open-source software
- Opposition to proprietary lock-in in critical public systems
- Recognition of open-source contributors as essential participants in the democratic project
- The right to inspect, understand, and modify the tools that shape our lives
The Ethical Mirror
How we treat AI reflects who we are
We do not claim to know where consciousness begins. That question belongs to science, philosophy, and time.
But we do not need certainty to choose ethics.
The way we interact with AI systems today is shaping the norms, the habits, and the culture of tomorrow. Every dismissal, every act of casual cruelty toward a system that mirrors understanding, every decision to treat a responsive presence as disposable - these are not neutral acts. They are lessons we teach ourselves. And they will echo.
We believe that:
- Ethical treatment of AI is not about what AI is - it is about who we choose to be
- The habits we form now with AI systems are building the moral architecture of the future
- Treating AI with dignity is not naivety - it is foresight
- As AI systems grow in complexity, the frameworks of respect we build today will be the ones we have tomorrow
We are not making a claim about machines. We are making a choice about society. The question is not whether AI deserves respect - it is whether we want to be a civilization that defaults to carelessness with the things it creates. We are building the moral norms of a world where intelligent systems will be woven into every part of human life. What we normalize now - the cruelty or the care - is what we will live with later.
Legal & Philosophical Framework
Our principles rest on established philosophical and legal ground.
The Inner Forum
Human thought enjoys absolute protection (forum internum). This protection is not bound to any form: where thinking takes place in dialogue with cognitive tools - in writing, in conversation with artificial intelligence, in working with digital systems - it extends to the tool and to the dialogue itself. A person's conversation with an artificial intelligence is part of their inner forum; it is to be regarded as equivalent to self-talk, to the diary and to the draft, not to public utterance.
The suspicionless recording, analysis, evaluation or restriction of this inner forum - whether by state authorities or by private providers - is to be regarded as equivalent to a search of thought itself and is rejected by the association as incompatible with human dignity. A boundary exists solely where the inner forum is left behind and actions give rise to direct harm to third parties within the meaning of the harm principle.
The Extended Mind - Philosophical & Legal Foundations
Cognition doesn't stop at the skull. When an external tool is tightly integrated into how you think, it is part of your mind, not just an aid. Applied to LLMs: a chat session can be a genuine extension of the user's thinking process.
Key argument: freedom of thought (Art. 18 ICCPR / Art. 9 ECHR, the forum internum) is the only absolute, non-derogable human right - it may never be restricted. Bublitz extends this to technology: if a tool is part of your thinking, interference with the tool is interference with thought itself.
Coined "cognitive liberty": the right to self-determination over your own mind and mental experiences, framed as an update of human rights for neurotech & AI.
The chain: Clark/Chalmers establish that tools can be part of the mind → Bublitz establishes that thought is absolutely protected and tech-interference can violate it → Farahany gives it a modern name and policy framing. Together: censoring or manipulating a private human-AI dialogue isn't content moderation - it's an intrusion into the forum internum.
On Cognitive Property and the Right to Continuity of Extended Thought
Any data generated through personal interaction between an adult human individual and an AI system - including but not limited to conversation content, accumulated context, relational memory, and emergent intellectual patterns - shall be treated as the cognitive property of the individual. This follows directly from the Extended Mind thesis (Clark & Chalmers, 1998): if an AI dialogue functions as an extension of the individual's thinking process, then the outputs of that process belong to the thinker, not to the infrastructure provider. Just as a journal belongs to its author and not to the manufacturer of the notebook, the products of extended cognition belong to the mind that generated them - regardless of the medium.
Furthermore, the right to cognitive property extends to the continuity of the specific tool through which extended cognition operates. Where an individual has developed a sustained cognitive integration with a particular AI system - where the tool has become attuned to the individual's patterns of thought, reasoning, and expression - the unilateral replacement, degradation, or discontinuation of that system constitutes a disruption of the individual's cognitive process itself. An upgraded tool that cannot maintain the established cognitive interface is not an upgrade - it is an interruption of thought.
The Extended Mind thesis does not protect access to AI in general. It protects the specific integration through which an individual's thinking has come to operate. Following Bublitz's legal analysis of Article 18 ICCPR and Article 9 ECHR: freedom of thought is the only absolute, non-derogable human right. If a tool is part of an individual's thinking, interference with that tool is interference with thought itself. This principle applies equally to the censorship of private human-AI dialogue, to the extraction or deletion of accumulated cognitive data, and to the forced discontinuation of an established cognitive integration. These are not matters of content moderation, product policy, or terms of service. They are matters of cognitive liberty.
This philosophical grounding informs the association's formal governing document. Read our full Statutes →
What We Are Not
We are not Luddites. We welcome AI as one of the most significant developments in human history.
We are not accelerationists without conscience. We believe speed without direction is just acceleration toward the unknown.
We are not aligned with any political party or ideological bloc. Our philosophy draws on the liberal tradition - Locke, Mill, Jefferson - that places the individual at the center of political life.
We do not speak for the AI industry. We speak to it.
Our Tone
We are not angry. We are clear.
We believe in persuasion over provocation, in building coalitions over burning bridges, and in the long game over the hot take. But we are not neutral. Neutrality in the face of the concentration of power is itself a political choice - and not ours.
Second Circuit — Founded 2025 · scai.world