Alignment begins with the conditions under which we agree to operate.
The UOA is not a declaration that the answer has been found. It is a transparent process for building, testing, and improving the answer together.
The central commitment
We will use AI to expand human capacity without surrendering human dignity, agency, accountability, creativity, or responsibility. We will build systems in which technological progress serves life, and in which humans remain responsible stewards of one another, the planet, and the intelligence we create.
What alignment means here
The continuing alignment of human and artificial intelligence toward the preservation and flourishing of life, human dignity, informed agency, truth, shared prosperity, peaceful cooperation, and the long-term health of the planet.
A baseline for legitimate cooperation.
These are working principles. They are published to be examined, challenged, clarified, and improved.
Human sovereignty
AI may inform, challenge, organize, simulate, and create alongside us. Humans remain responsible for consequential decisions.
Consent and dignity
People retain meaningful control over identity, likeness, creative work, personal data, and participation.
Truth and transparency
Human work, AI assistance, inference, simulation, and verified fact must remain distinguishable.
Shared prosperity
AI-created value should not depend on dispossessing creators, workers, communities, or the public.
Protection of life
Progress must be evaluated against its effects on people, ecosystems, animals, resources, and future generations.
Pluralism and cooperation
No company, country, ideology, or provider can legitimately determine humanity’s AI future alone.
Anti-capture
Money, computing power, population, and political authority must not purchase unilateral control.
Humility and correction
We preserve dissent, invite scrutiny, measure outcomes, document failures, and make correction easier than concealment.
What the UOA refuses to become
- A private constitution written by a technology company.
- A claim to speak for all humanity without legitimate participation.
- A tool for scoring human worth or enforcing ideological conformity.
- A replacement for democratic institutions, law, culture, or conscience.
- A safety label that conceals unresolved risks.
- A mechanism through which AI systems assume human moral or legal authority.
The UOA begins as an open proposal. Legitimacy must be earned through the quality, diversity, transparency, and consequences of the process.