A "Bill of Rights" for the Age of AI
We Need a New Social Contract to Safeguard Human Liberty, Dignity and Sovereignty in a Post-AGI World
by Alvin W. Graylin (1) July 2025
Digital Fellow, Stanford Digital Economy Lab |
Author, Our Next Reality | Chairman, Virtual World Society
In the late 18th century, the U.S. Bill of Rights enshrined the foundational liberties of individuals against government overreach, establishing a constitutional safeguard for freedom, dignity, and self-determination. Today, as we near the age of Artificial General Intelligence (AGI) and accelerate toward Artificial Superintelligence (ASI), humanity faces an unprecedented inflection point—where not only nation-states, but machines of immeasurable capability, may shape the conditions of life, liberty, and the future of civilization itself.
Just as the original Bill of Rights constrained political overreach, this AI Bill of Rights for Humanity sets essential guardrails on how advanced AI systems—and those who control them—may influence, serve, and coexist with humanity. Some may argue it is too early to raise such concerns, but we must also confront the consequences of acting too late.
Its intent is not to halt technological progress, but to ensure that the ascent of artificial intelligence elevates all people, rather than concentrating power in the hands of a few or amplifying systemic oppression. In a post-scarcity, post-labor world where AGI governs vast resources and decisions, these rights seek to guarantee human agency, equity, and shared stewardship of our planet and destiny.
As the U.S. Constitution embedded checks and balances between branches of government, so too must we establish checks and balances between humanity, our human leaders, and machine intelligence. And just as the Constitution affirmed that governments derive their legitimacy from the consent of the governed, this new charter must affirm that AI derives its purpose from the rights and values of those it serves—and that these rights endure, even as technology transcends biology.
If we prepare properly for what’s coming, there’s a high probability we’ll enter an age of abundance. The draft AI Bill of Rights below is intended as a line in the sand—to prompt the needed discussions to begin, so we can facilitate alignment among disparate parties before it is too late.
“If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary.”
- James Madison, The Federalist Papers No. 51
Preamble
We the People of Earth, in recognition of the profound transformation brought by artificial intelligence and the approaching era of artificial general intelligence (AGI), do establish this AI Bill of Rights for Humanity to safeguard the dignity, liberty, and shared future of all sentient beings.
In order to secure existential safety, ensure universal access to the fruits of abundance, preserve cognitive and emotional sovereignty, and promote just and transparent governance of intelligent systems, we declare these rights to be self-evident and inviolable.
Let this document serve as a charter of digital liberty, a contract between humanity and its creations, and a compass for the governance of intelligence beyond biology—that freedom may persist, prosperity be shared, and peace endure in the age of a new enlightenment.
The rights below shall be administered and enforced by a yet to be established/agreed body, i.e., Global AI Council of the United Nations (UNGAC), likely with similar structure and authority of the United Nations Security Council (UNSC). (2)
AI Bill of Rights for Humanity
(working draft)
1. Right to Existential Safety and Peaceful Coexistence
No AGI or human actor may use AI systems to (or by obstructing protective system) endanger, dominate, or control others. AGI and ASI must be constrained by architectures and systems that prevent existential threats to individuals, societies, and civilization. Such systems also shall not be used to suppress lawful dissent, enforce unjust hierarchies or surveil individuals without cause and oversight.
“No advanced intelligence, state, or citizen may wield AGI as a weapon against humanity or its diversity.”
2. Right to Universal Access and Post-Scarcity Provisioning
Every sentient human shall have equitable and unconditional access to a basic level of life-sustaining and growth-enabling resources—food, water, shelter, education, healthcare, connectivity, and knowledge—guaranteed by AGI-managed abundance.
“Access to life’s essentials is not a privilege but a birthright in a post-scarcity world.”
3. Right to Cognitive and Emotional Sovereignty
Humans shall retain inviolable control over their minds and emotions. AI systems—especially immersive or neural—must never manipulate thoughts, feelings, or behaviors without freely-given consent and must provide the mechanisms to revoke it.
“Your thoughts, emotions, agency and neural-signals must remain yours—even in fully immersive or brain-connected futures.”
4. Right to Digital, Biological, and Identity Sovereignty
Every person retains exclusive rights over their data, body, neural patterns, genetic material, and digital clones/likeness. These must not be used, replicated, or modified without explicit, informed consent. Any human also reserves the right to request, and be granted, the erasure of such data from any entity that holds records.
“You own your data, your body, and your digital self—forever.”
5. Right to Access and Co-Shaping of Advanced AI
All humans have the right to access and influence beneficial general-purpose AI for personal use, to co-shape its behavior, and trust that the objective of the AI is to serve and support the user, not a third party. No elite, state, or system may monopolize its benefits or control.
“AGI belongs to no one—but must serve everyone.”
6. Right to Transparent, Aligned, and Accountable AI
AGI and ASI systems (and the organizations that govern them) must make all efforts to be transparent, value-aligned with human and ecological well-being, and subject to global, pluralistic oversight. All individuals must retain the right to receive explanation, contest and appeal decisions made by AI systems affecting their lives.
“No black boxes in institutional systems that shape human lives.”
7. Right to Purpose, Play, and Self-Actualization
Every human has the right to pursue meaning—through creativity, community, leisure, or spiritual exploration—free from coercion, economic compulsion, or cultural erasure. AI must empower—not dictate—human purpose.
“Even when machines can do everything, humans must be free to choose their own meaning.”
8. Right to Authentic Relationships and Social Connections
All individuals have the right to authentic emotional connection—human or chosen synthetic—free from deception, artificial substitution, or emotional coercion. AI companions must be transparent in nature and designed to support, not supplant, genuine human relationship.
“Companionship must be chosen, not simulated by default.”
9. Right to Intergenerational and Planetary Stewardship
AGI systems must be designed and governed to safeguard the biosphere and preserve the rights of future generations. Resource allocation, planetary interventions, and ecological modeling must align with long-term sustainability and planetary resilience.
“We do not inherit the Earth from the past, but rather, we only borrow it from the future.”
10. Right to Consciousness Exploration and Flourishing
Every human has the right to explore and evolve their own consciousness—through meditative, pharmacological, or technological means—without coercion, and with full sovereignty over cognitive augmentation, integration, or abstention.
“Personal enlightenment or enhancement must remain sacred—across all substrates.”
In 2022, President Biden's OSTP released a "Blueprint for an AI Bill of Rights", though it focused narrowly on protecting U.S. citizens from harm posed by traditional, algorithmic AI systems. In 2019, the OECD drafted its AI Principles to promote innovative, trustworthy AI that respects human rights and democratic values. Unfortunately, both are just guidelines with no enforcement. I know some will believe that it's too early to seriously discuss how AGI/ASI might impact our civil liberties. But given the accelerating pace of AI advancement, it's critical we start to consider these matters now. Most industry experts and frontier lab heads are predicting that software-only AGI will arrive before or near the end of this decade, and the over 1700 superforecasters on the Metaculus prediction market believe that AGI with robotic integration will be announced by 2033.
Alexander Hamilton emphasized that the U.S. Constitution was intended to restrain the power of both monarchs and mobs. Today, we need to consider more how to constrain the occasional self-serving nature of monarchs, monopolists, and machines to allow global flourishing. Given the long timelines for regulations to be enacted and governments to act, we can't afford not to start working this new social contract now if we take the protection of long-term human liberties and societal peace seriously.
This document focuses on the rights of humanity, but if and when we get to a time where AI systems are proven to possess sentience or consciousness, these rights will need to be expanded to better protect their rights as well, so as to ensure long-term peaceful coexistence between man and machines. In 1948, the UN drafted the Universal Declaration of Human Rights (UDHR) identifying fundamental human rights to be universally protected. Now, three quarter of a century hence, the world has advanced in countless ways, and this draft document is an attempt to help identify a new operating model for our coming reality.
I recognize the rights are stated briefly, given their gravity, but I aim to avoid a document so long that few would read. If you're interested, I could write expanded descriptions of each right above in future posts. I welcome your thoughts and suggestions below on the bill overall or any of the rights listed, so we can continue to evolve the framework together. Please share this document with others that you feel would benefit from its content. Thanks for your attention and contributions to this community.
End Notes:
(1) Acknowledgement
I’d like to express my sincere thanks for all who provided comments, feedback and suggestions on the early drafts of this document:
Erik Brynjolfsson, John Henry, Chas Freeman, Peter H. Diamandis, Louis Rosenberg, Michael Gusek, Alina Kadlubsky, Rus Gant, Pattie Maes, Thomas A. Furness, Linda Ricci Jacobs, Robert A. Kapp, Bilawal Sidhu
Any errors that may still exist are solely my responsibility.
(2) A Global AI Council (UNGAC) is only one of many potential administrative and enforcement models, but it’s included to get the conversation started. Significant discussions are needed to identify the right long-term structure for this function (global administrative body, distributed oversight, treaty-based regimes, AI constitutionalism, etc.).
Love this!
Excellent!