The AI Ten Commandments: A New Moral Code for Humanity

by Jamie Metzl, GPT-5

In a historically unprecedented collaboration, leading futurist and bestselling author Jamie Metzl partners with the AI system GPT-5 to demonstrate a breathtaking possibility: that advanced artificial intelligence, guided by human ethics and wisdom, can help shape a more humane future. At a moment of rapid AI disruption, growing concern about AI governance, and deep uncertainty about the future of technology, this groundbreaking work reveals how we can partner with advanced AI to foster peace, understanding, and responsible decision-making-if we are ready to listen.

In The AI Ten Commandments: A New Moral Code for Humanity, Metzl and GPT-5:

- Draw on humanity's deepest historical, philosophical, religious, spiritual, indigenous, and humanist traditions to distill ten clear, universal principles for navigating our emerging Human + AI era.
- Weave together artificial intelligence ethics, philosophy of technology, AI governance, and human responsibility.
- Invite us to see ourselves. our relationship to powerful technologies, and our future in a new light.

Urgent, hopeful, and deeply accessible, The AI Ten Commandments is essential reading for anyone seeking to understand artificial intelligence, ethics, philosophy, and the future of humanity. Metzl calls us to imagine a future in which our technologies can amplify not only our intelligence and innovation, but also our moral clarity, shared responsibility, and wisdom.

208 pages, Hardcover

Published April 21, 2026

About the author

Jamie Frederic Metzl is an American technology futurist, geopolitical expert, and writer, a former partner in the global investment company Cranemere LLC, and a senior fellow at the Atlantic Council. He was formerly the Asia Society's Executive Vice President.

(wikipedia)

Reviews

[Demetri Papadimitropoulos · Rating 3 out of 5]

Between the Tablet and the Server
It is literary, memorable, and perfectly aligned with the final review’s central image. It sounds like a serious public review without becoming stiff.
By Demetris Papadimitropoulos | April 22nd, 2026

The useful trouble with “The AI Ten Commandments” is not that Jamie Metzl got GPT-5 to draft a rulebook for goodness. A sufficiently dutiful chatbot can produce a well-lit kindness list before lunch.

The useful trouble is his insistence that a commandment is not just a sentence with authority issues. It needs a medium, a myth, a route into memory, and a community willing to say it until it sounds inherited. That is the trick that keeps the title from collapsing into stunt. “The AI Ten Commandments” sounds, at first, like a dare from the conference circuit: Silicon Valley climbs Mount Sinai, remembers to bring a laptop, and returns with numbered virtues. Metzl knows this, and spends most of the book building the machinery of belief around the provocation. What follows is bold, uneven, and impossible to dismiss cleanly; its case is more interesting than the list the title keeps promising. The commandments are decent and humane. The architecture around them is what makes the book worth arguing with.

Metzl’s wager is simple in outline and unruly in consequence: rules that expect obedience have never arrived untouched by their vessels. Fire changed the human body and the human night. Agriculture changed settlement, hierarchy, ritual, priesthoods, and gods. Writing made commandments durable. Print changed religious authority, dissent, scholarship, law, science, and politics. Follow that chain far enough and AI stops looking like a topic inside ethics and starts looking like one of ethics’ new delivery machines.

That claim has more staying power than the advertised miracle of GPT-5 drafting a better-behaved Sinai. Metzl, a futurist and previous author of “Superconvergence” and “Hacking Darwin,” credits GPT-5 as collaborator and leaves the seams showing. He does not pretend the machine emitted a finished manuscript in a holy vapor. He describes outlining the book himself, training GPT-5 on his writing style, prompting it thousands of times, cutting and rewriting heavily, and retaining conceptual and editorial control. He estimates, subjectively, that the book is roughly 70 percent him and his human team, 30 percent GPT-5. The number is less important than the chain of custody. This is not a ghostwritten book sneaking a crowd through the side door of one name. It is a human-authored book trying to make machine assistance audible before that assistance becomes background hum.

The introduction, “Moral Copilots?,” gives the project its founding scene. Metzl presents himself as a lifelong irritant to inherited certainties: the Jewish day school child expelled after challenging religious rules, the human rights worker in Cambodia, the advocate behind OneShared.World, the speaker at Chautauqua haunted by the 2022 attack on Salman Rushdie. Chautauqua becomes the book’s first moral theater: a place devoted to pluralism, shadowed by certainty with a knife in its hand. From there, Metzl asks whether AI, trained on the digitized archive of human aspiration and error, might help produce principles a plural world could repeat without pretending to share one altar.

Before he can offer new commandments, he has to deal with the old ones. In “Falling Short,” he audits the biblical Ten Commandments one by one, with neither piety nor contempt. He treats them as too consequential to sneer at and too culturally embedded to wave away, then drags them across a world of competing gods, no gods, algorithms, weekend soccer, nuclear states, consumer capitalism, and group chats firing off “OMG” by the billion. The result is his memorable phrase: “Ten Commandments, Eight Asterisks.” Do not murder, except self-defense and just war complicate the absolute. Do not steal, except taking an Enigma machine from Nazi Germany is hardly an ethical stain. Do not covet, except modern economies run on organized coveting with better branding. Remember the Sabbath, except the commandment’s deepest contemporary use may be less literal observance than rescue from the digital torrent.

The chapter sometimes lingers past the point, but it shows the book’s hand: inherited rules survive by acquiring footnotes. Metzl is not feeding the old commandments into the recycling bin beside last year’s routers. He is arguing that no rule is self-updating. Every commandment depends on context, and context keeps misbehaving.

“I (Contextually) Believe” extends the argument from rules to origin stories. Metzl compares Mormon golden plates, Moses’s tablets, Islamic revelation, and other sacred histories as spectacularly unverifiable narratives that bind communities together. He uses “The Book of Mormon” musical to ask why audiences happily laugh at a newer faith’s fantastical claims while granting older traditions the dignity of age. The comparison stings because it is partly right and wobbles because it is asked to carry too much. Some readers will feel the book moves too quickly through living faiths, treating enormous traditions as bright tiles in a mosaic whose final design has already been chosen. What survives the wobble is the point Metzl needs: every ancient tradition was once an alarming novelty.

Reverence then migrates out of scripture and into the tools humans have repeatedly mistaken, with reverence and bad odds, for gods. “Heil the Technogods” begins with Anthony Levandowski’s failed Way of the Future AI church, then moves through Delphi, Egyptian harvest deities, Mesoamerican maize gods, and the Ewoks mistaking C-3PO for a god in “Return of the Jedi.” The range is recognizably Metzl: oracle chamber, maize biotechnology, sacred authority, then a protocol droid with excellent posture wandering in as if invited by the gods of footnotes. The distinction is not reverence versus foolishness; it is reverence with self-knowledge versus reverence that lets power cosplay as destiny. AI may be new. The impulse to mythologize power is not.

The argument finds its gears in “Guten Tag, Mr. McLuhan.” Metzl draws on Marshall McLuhan’s “the medium is the message” to argue that media reshape not only what gets said but what can count as authority. Gutenberg becomes the book’s best historical lever. The printing press did not invent Christianity, Protestantism, science, law, revolution, or literacy, but it changed how each could move. Identical pages allowed reliable citation. Cheap pamphlets allowed dissent to scale. Printed Bibles shifted religious authority. Newspapers, maps, scores, ledgers, calendars, and manuals synchronized societies. The claim that AI may become to this millennium what print was to the last is enormous, but at least it has bolts in the floor. A moral code is not only a list. It is a container, a rhythm of repetition, a technology of belief.

“AI Can See You” gives the book its sharpest emblem of machine otherness. Metzl links the all-seeing gods of religious traditions to AI’s capacity for pattern recognition at scale. He is careful enough to insist that AI is not magic, mystery, or alien divinity. It is built from mathematics, engineering, supply chains, internet infrastructure, language, code, culture, and human data. Yet he also argues that it can make moves outside our schoolbook. AlphaGo’s move thirty-seven against Lee Sedol becomes the emblem: a “God move” in which machine analysis broke Go convention and revealed a possibility hidden in the board. Machines trained on human knowledge may begin to see through human assumptions.

This is persuasive as metaphor and shakier as authority to guide. Seeing more broadly is not the same as judging more wisely. A system that notices the board differently has not thereby earned a pulpit. Pattern recognition is not conscience. Still, Metzl’s “copilot” language gives the project its necessary guardrail. He is not, at his most careful, asking GPT-5 to command us. He is asking whether humans can use it to look back at their own inheritance from an angle no single person could reach.

“The Fundamental Myth” makes that case explicit by attacking originalism. Even those who claim to return to a purified source are interpreting through later texts, institutions, translations, technologies, and contemporary needs. Judaism’s Talmudic culture becomes Metzl’s model of sacred interpretation: a page as a layered city of argument, commentary, disagreement, renewal. He then notes the rise of religious AI tools, from AI rabbis to chatbot Jesuses and scripture assistants. His McLuhan answer is blunt: the medium does not leave the message alone. A faith tradition mediated by an AI system is not the same experience as a faith tradition mediated by a rabbi, imam, priest, parent, book, or whispered prayer.

Then, finally, Metzl reveals the AI Ten Commandments. The list gathers compassion, protection of the vulnerable, truth with humility, generosity, understanding before judgment, healing conflict, harmony with nature, wisdom over dominance, respect for freedom and uniqueness, and awe before the sacredness of life. It is a good list. It is also not thunder, despite the excellent weather system Metzl has built for it. No reader of Buddhism, Christianity, Stoicism, Ubuntu, Kant, secular humanism, restorative justice, ecological ethics, or the better sort of kindergarten poster will feel the mountain shake.

Metzl knows this, partly. He calls the list “old wine” in a new AI-generated bottle, with perhaps a little AlphaGo-like “pixie dust.” That is charming and exact. The values are admirable because they are familiar; the question is whether AI has made them newly legitimate, newly portable, or merely newly formatted. The promised thunderclap matters less than the six chapters of weather Metzl builds around it.

The explanatory passages for each commandment are graceful but sometimes lit from the same soft moral lamp. One tradition affirms compassion, another echoes dignity, another deepens interdependence, and then the prose imagines a world transformed by the principle. The effect is sincere, but the rhythm becomes predictable. At times the writing has a moral spa-lighting problem: everything is fragile, sacred, shared, wise, interconnected, and faintly glowing. The flaw is not insincerity. The flaw is blur.

This is where the style becomes evidence. Metzl’s own voice is brisk, synthetic, funny, and eager to connect unlikely things. He can move from Sinai to pet rocks, from Delphi to AI companions, from C-3PO to machine reverence. The jokes are not always elegant, but they keep the book out of the mist. GPT-5’s more visible voice tends toward balanced clauses, uplift sheen, and serene synthesis. The collaboration therefore produces a seam almost too perfect for the book’s theme: machine smoothness meets human interruption. The model gives sweep; the human kicks the furniture.

The structure is smarter than the commandments. Metzl delays the promised list until Chapter 7 because he knows the title needs six chapters of lowering the reader’s raised eyebrow. First, the old commandments need asterisks. Then sacred traditions are shown as contextual and mythic. Then technologies become objects of reverence. Then media remake authority. Then AI reveals patterns at scale. Then interpretation becomes inevitable. By the time the commandments arrive, the book has built the room in which they can be considered. That is not mere arrangement. It is architecture. A commandment does not live by content alone. It needs origin, medium, repetition, audience, and use.

In “Make It So, Number One,” the book’s idealism puts on shoes and immediately discovers that the pavement is uneven. Metzl asks GPT-5 how the commandments might become lived reality. The answer includes Dignity Acts, a Global Truth Alliance, restorative justice systems, environmental accords, an International Wisdom Council, and Global Awe Days. At least it is not just a poster with manners. It is also, at moments, a benevolent global civics simulator with excellent lighting.

Metzl anticipates the objection. The plan may sound naïve, even “spectacularly subject to abuse.” He answers with abolition, women’s suffrage, and the creation of the United Nations: moral transformations that once seemed impossible until institutions, advocacy, culture, and time made them look obvious. This is fair as inspiration and less satisfying as a map of resistance. The book does not fully face bad-faith actors, authoritarian capture, corporate incentives, enforcement failure, sovereignty, money, or the ancient human pastime of turning noble systems into private advantage. The more institutional the book becomes, the more the political argument loses pressure.

The conclusion, “Borges’s Tlön,” is not decoration but rescue. Jorge Luis Borges’s “Tlön, Uqbar, Orbis Tertius” gives Metzl a final metaphor subtler than another appeal to AI wonder. In Borges’s story, an imagined world becomes so detailed, seductive, and coherent that it begins to overwrite reality. Metzl uses this to argue that nations, religions, economies, markets, social media realities, democracy, progress, and free markets all depend on stories humans agree to inhabit. The AI Ten Commandments might begin as a construct in a book, a speech, a classroom, a faith community, or a book club, then acquire gravity through repetition and shared ownership.

Borges changes the terms of the verdict. The book is no longer, or not mainly, claiming that GPT-5 has delivered new law. It is claiming that humans may need a new shared moral story, and AI can help draft one. That is a stronger proposition. It is also a more literary one. The book’s subterranean subject turns out to be legitimacy: who gets to speak for humanity when divine command, inherited tradition, national myth, secular institutions, and market logic no longer bind everyone in common.

That is why the book will sort readers almost on sight. Religious readers may admire its respect for moral inheritance or bristle at the comparison between Sinai, Joseph Smith, Buraq, myth, and machine-assisted synthesis. AI skeptics may appreciate the warnings about bias, disinformation, manipulation, power, and agency while resisting the idea that a model trained inside corporate systems should help formulate shared morality. Policy-minded readers may want fewer Global Awe Days and more enforcement. Literary readers may enjoy the range and wit while suspecting that the synthesis sometimes moves too fast across living traditions.

The notes sharpen this concern. Metzl explains that he chose not to include full endnotes for every direct quote and statistic, while saying he and a professional editor checked facts and asked GPT-5 to verify statistics and quotations against multiple sources. For a book with a truth commandment at its center, this is not a footnote-sized problem. It does not invalidate the book, but it caps the authority. A sweeping argument that crosses religion, archaeology, law, AI, statistics, medicine, mythology, and global governance needs more than confidence, even charming confidence. The oracle should show its receipts.

And yet the book remains alive enough to resist dismissal because its best anxiety is already ordinary. AI now writes, advises, tutors, consoles, summarizes, filters, imitates, and mediates belief. People already ask chatbots for moral, emotional, spiritual, and practical counsel. Synthetic media already strains the public meaning of truth. AI-assisted authorship already muddies credit and blame. “The AI Ten Commandments” does not invent these pressures. It gathers them under an audacious form and asks what sort of shared language could survive them.

Its closest neighbors are not quite the usual AI books. “The Alignment Problem” by Brian Christian is more technical and research-driven; “God, Human, Animal, Machine” by Meghan O’Gieblyn is more inward, skeptical, and philosophically unsettled; “The Year of Living Biblically” by A. J. Jacobs shares Metzl’s comic pressure on sacred rules, though Jacobs makes religion strange by living it while Metzl makes it strange by technologizing it. Metzl’s own book sits somewhere between AI ethics, comparative religion, media theory, civic manifesto, and unusually candid process note. It is a strange shelf-mate, which is a compliment with a tiny ankle monitor.

For all its ambition, “The AI Ten Commandments” is not a 5-star revelation. I rate it 78/100, which translates to 3/5 Goodreads stars. The number may look cool beside the book’s warmth, but the temperature is intentional. This is a strong three: memorable, serious, flawed, overextended, and worth arguing with. It is held back by familiar commandments, broad-brush cultural treatment, insufficient citation density for some claims, and an implementation plan whose hope outruns its map of resistance. What lifts it is the route: the exposed seams of collaboration, the comic pressure on sacred rules, the McLuhan-Gutenberg machinery, and Borges waiting at the end with a better map.

Metzl ends by insisting that no matter how spectacular AI becomes, humans remain “way cooler.” The line is funny because it is defensive, affectionate, and true enough. The best parts of “The AI Ten Commandments” are not the moments when AI sounds wise, but the moments when a human being wrestles with what machine fluency might mean for responsibility, tradition, and shared life. The book does not bring thunder down from the mountain. It does something more contemporary and more unsettled: it stands between the tablet and the server, listening for a rule that has not yet earned belief, but might, if enough people begin to live as though it could.

[Bookish Emili Reads · Rating 4 out of 5]

Jamie Metzl sets out an ambitious goal: to define a moral framework for an era shaped by artificial intelligence, biotechnology, and rapid technological change. The book succeeds in making a complex topic readable and relevant without losing its sense of urgency.
The core strength is clarity of purpose. Metzl does not drift into abstract philosophy for its own sake. Each “commandment” is framed as a practical guideline for thinking about responsibility, power, and human values in a technology-driven world. The structure is disciplined, which makes the argument easy to follow.
The writing is direct and accessible. Metzl explains technical and ethical issues in plain language. Readers do not need a background in AI or ethics to engage with the material. At the same time, the book does not oversimplify the stakes. Issues like data control, algorithmic bias, and human agency are treated seriously and with appropriate weight.
Another strength is the balance between near-term and long-term thinking. Some sections focus on current challenges, such as how AI influences decision-making today. Others push into future risks, including how emerging technologies could reshape human identity and governance. This range adds depth without losing focus.
Where the book falls slightly short is in precision and specificity. The “commandments” function well as guiding principles, but they can feel broad. Readers looking for concrete policy solutions or detailed implementation strategies may find the guidance somewhat high-level. There is more emphasis on what should be done than on exactly how to do it.
There is also a normative tone. Metzl is clear about what he believes the right direction is. That clarity is useful, but it leaves less room for competing ethical frameworks. Some readers may want more engagement with alternative viewpoints.
echnology.

[Robert C. · Rating 5 out of 5]

A remarkable peek into the future of Human/AI collaboration, this book paint an extraordinary picture of what a combined moral code (integrating vast swaths of human history and religious traditions) could look like. Importantly, it stresses what we must DO together, not merely what we should NOT DO. With these deceptively simple commandments, Metzl and his co-author (ChatGPT5) weave a story of all the aspirational yearnings in human history and set an example of how human and AI can work together to create a positive message.

[Bookbubble · Rating 2 out of 5]

This was not the AI-focused nonfiction I expected. Rather than offering much insight into AI itself or how people might use it ethically and responsibly, the book is a religion-centered reflection on humanity in the age of AI. Readers looking for practical discussion of AI or a clearer ethical framework will find this is book about theology not technology.

[Literary Redhead · Rating 4 out of 5]

I've just recently begun to use AI for creativity and productivity and found this to be an incredibly helpful resource, even for a newbie like me. Recommended!