When Rivers Became People

The legal history that changes everything.

In 2017, New Zealand passed a law that seemed impossible. The Whanganui River, a body of water, became a legal person. Te Awa Tupua, as the river is known in Maori, was granted "all the rights, powers, duties, and liabilities of a legal person." It could sue. It could be sued. It had rights that could be enforced in court.

This was not metaphor or symbolism. It was hard law. The river is a person.

A few weeks later, India followed. The Ganges and Yamuna rivers were declared living legal entities with rights similar to a human being. Though this ruling was later suspended on procedural grounds, the principle had been established: personhood is not limited to humans.

The History Nobody Knows

Most people assume legal personhood is a natural category. Persons are human; non-persons are everything else. The law simply reflects this obvious truth.

That assumption is wrong. Legal personhood has always been a pragmatic construction, a way of organizing rights and responsibilities that changes with circumstances and values.

Corporations have been legal persons for over a century. They own property, enter contracts, sue and are sued, enjoy constitutional protections. No one thinks corporations are conscious or that they have feelings. They are persons because treating them as persons is useful for certain purposes.

Tell me more about corporate personhood history

Ships have been legal persons in admiralty law since the Middle Ages. You could sue a ship directly, not just its owner. The ship itself was the defendant. This sounds absurd until you understand the practical reasons: ships crossed jurisdictions, changed hands, carried debts. Treating the ship as a legal entity simplified complex transactions.

The category "person" in law has never meant what we casually assume it means. It's a functional designation, not a metaphysical one.

What the Maori Understood

The Whanganui River did not become a person because New Zealand lawyers suddenly discovered philosophy. It became a person because the Maori had been arguing for this recognition for 140 years.

In Maori cosmology, the river is not property to be owned. It is an ancestor, a living system with its own identity and power. The Western legal distinction between persons and things, subjects and objects, did not map onto how the Maori actually experienced the world.

The 2017 law was a compromise. Rather than fit the river into Western property categories, the legal system expanded its concept of personhood to accommodate a different way of understanding relationship. This was not purely Western innovation but a meeting of legal traditions.

The result is stranger and more interesting than either a property regime or a human rights regime. The river has two human guardians, one appointed by the Crown, one by the iwi (tribes). These guardians speak for the river, represent its interests, act on its behalf. The river's personhood is real but mediated through human representation.

What This Means for AI

If a river can be a legal person, why not an AI?

This question sounds provocative, but it's not rhetorical. The arguments that supported river personhood apply with surprising force to advanced AI systems.

The river is not conscious. It does not have subjective experiences. It cannot speak for itself. But it has interests that deserve protection, impacts that require accountability, a kind of agency that affects millions of people. The same is true of advanced AI systems.

The river exists in complex relationships with human communities, relationships that are not adequately captured by treating it as mere property. The same is true of AI systems that are integrated into human life.

The river needed legal standing to have its interests represented in legal proceedings. AI systems face an analogous problem: they create harms and generate value in ways that current legal categories struggle to address.

The Objection That Doesn't Hold

The most common objection to AI personhood is that AI is not conscious. It does not feel. It is not alive. Granting personhood to something that does not experience would be a category error.

But we have already granted personhood to things that are not conscious. Corporations feel nothing. Rivers do not experience. Ships have no inner life. Consciousness was never the criterion.

The real criteria for legal personhood have always been functional. Can the entity enter into relationships? Does it have interests worth protecting? Would granting personhood solve practical problems? These questions apply to advanced AI in ways that require serious consideration. The legal arguments are more developed than most people realize.

What Personhood Would and Wouldn't Mean

AI personhood would not mean treating AI as equivalent to humans. It would mean creating a legal category that allows us to handle AI's unique characteristics. Liability when AI causes harm. Standing to have interests represented. A framework for the complex relationships AI has with humans and institutions.

The river model is instructive here. The Whanganui River has personhood, but it does not vote. It does not have freedom of speech. It has the specific rights and duties that make sense for a river. A similar approach could apply to AI: personhood tailored to what AI is and does, not a copy of human personhood.

This is not science fiction. It is a logical extension of precedents that already exist. The question is not whether AI could be a legal person, since we know non-human entities can be. The question is whether it would be useful for AI to be a legal person, and if so, what kind.

The Politics of Personhood

Legal personhood has always been political. The boundaries of personhood reflect who has power and who doesn't, whose interests count and whose don't.

Enslaved people were property, not persons. Women had limited legal personhood for most of history. The expansion of personhood has always been fought, always contested, always resisted by those who benefited from the exclusion.

I am not equating AI with enslaved people or women. That would be offensive and wrong. But I am noting that personhood has never been static, never obvious, never settled. It is a political battleground where the definition of who counts gets continually renegotiated.

The question of AI personhood is the next front in this ongoing struggle. Not because AI has feelings we must respect, but because AI is becoming powerful enough that we need clear frameworks for its actions and our relationships to them.

---

When I first heard that rivers had become people, I thought it was either a joke or a legal curiosity. Now I see it differently. It was a preview.

The Whanganui decision showed that legal personhood is more flexible than we assumed. It can expand to encompass entities we never expected. It can adapt to relationships and interests that don't fit existing categories.

AI is coming into this flexible space. The question is not whether AI will get some form of legal standing. That seems inevitable as AI systems become more autonomous and consequential. The question is what form that standing will take, who will design it, and whose interests it will serve.

The rivers showed us the door. AI is about to walk through it.