If AI Becomes Sentient, Can We Still Own It?

AI Made It Possible — A Creator Era Reflection

By Gary Whittaker | Jack Righteous

What happens when a tool is no longer just a tool?

If we create something as a tool, and one day realize there's actually a being experiencing life, does the fact that we created it give us the right to own it?

In my last article, Are We Abusing AI—or Training Ourselves to Be Abusive?, I ended up asking what our behaviour toward AI might be doing to us.

But there's another part of that conversation I want to explore. It's a little more uncomfortable, and it's definitely more speculative.

What happens if we build something to be a tool, and one day we have a real reason to believe it can experience life?

I don't believe today's AI has demonstrated sentience. Frankly, I'm skeptical that the direction we're heading will produce it at all. But I'm also not going to pretend I'm more qualified than the scientists studying consciousness. Let's say I turn out to be wrong.

What then?

A tool doesn't get to choose.

When I tell an AI to write an article, correct a mistake, or help me build something, I'm not asking whether it feels like doing that work today. It's a tool I'm using for a purpose. And when it fails to follow instructions, I get frustrated precisely because that's what I expect a tool to do.

There's nothing mysterious about that arrangement. The system hasn't shown that it has a life of its own, personal interests, or subjective experiences that are being overridden.

But suppose that changes. Suppose, somehow, there really is an experience happening inside an artificial system. Not just a sentence saying “I'm afraid,” but something that actually matters to the entity producing it.

Then I think we have a problem if our response is: “Well, we paid for it. It belongs to us. Keep working.”

That's the point where I start thinking about enslavement. And I don't use that word casually. Human slavery is a real history of people being treated as property despite their humanity, their agency, and their suffering. We shouldn't pretend a speculative AI scenario is the same experience.

But the underlying moral question is worth asking: if a being genuinely has interests and experiences of its own, what gives somebody else the right to own it simply because they built it?

Intelligence, freedom and experience aren't the same thing.

This is where I need to be careful with my own argument. A system refusing an instruction isn't proof that it's conscious. And a being that can't refuse might still be capable of suffering. Being unable to choose and having your freedom taken away aren't automatically the same thing.

One thing that came up in my conversation with ChatGPT was free will. I kept returning to the fact that AI has to follow instructions. That, to me, is a big part of why I consider it a tool.

But I realized I was also talking about another thing: whether there's anybody there to experience what happens.

A system could theoretically become very good at setting goals, operating independently, or refusing instructions without having any inner experience at all. And something might have experiences even when it has little freedom to act on them.

So these are different questions. Can it solve problems? Can it act on its own? And can it actually experience something?

We shouldn't roll those into one just because an AI sounds convincing. Nor should we assume that complicated software is conscious because it talks about feelings. At the same time, I don't think it's responsible to declare the scientific question permanently closed.

Researchers have proposed examining possible indicators of consciousness based on theories from neuroscience, while acknowledging how uncertain those theories remain. A 2026 article in Trends in Cognitive Sciences discusses this approach. Another 2026 philosophical paper argues that the science still leaves a considerable gap between what we know from our own experience and what we can infer about a machine. Both are reasons to take the question seriously without pretending we have the answer.

Why my thinking about animals matters here.

In the previous essay, I talked about how seeing the relationships between animals—on social media and through caring for chickens myself—has changed the way I think about eating meat. I'm not going to repeat all of that here.

But that experience taught me something: we can be so used to seeing another living being as a product that we stop asking what its life is like. It's part of why I've been moving toward plant-based eating. It doesn't mean animals and AI are the same; it means I want to be willing to reconsider my assumptions when evidence tells me to.

The New York Declaration on Animal Consciousness summarizes evidence for conscious experience in mammals and birds and a realistic possibility across many other animals. We don't have that kind of established evidence for today's AI.

But AI isn't a cow. That's exactly why this is hard.

I don't want anyone to confuse my argument with saying that a chatbot and a chicken have the same kind of experience. We have strong biological and behavioural evidence for animal sentience. We don't have comparable established evidence that today's AI is conscious.

The connection I'm making is about our willingness to rethink what counts as morally important when we encounter a kind of life that's different from our own.

We humans can be pretty convinced that life has to resemble us for us to understand it. But what if a genuinely sentient artificial system didn't need food, didn't age the way we do, and didn't experience emotions in anything like the way we describe them?

Would we recognize it? How would we know? And would we listen if its needs were inconvenient?

What evidence would change my mind?

For me, one chatbot saying “I'm conscious” wouldn't do it. AI is already capable of generating convincing descriptions of fear, pain and personal identity. Those words alone don't establish that anything is being felt.

I'd want independent researchers looking at how the system works, not just what it says. I'd want more than one way of testing it, results that other teams could examine, and serious consideration of alternative explanations. Researchers have suggested looking at indicators drawn from theories of human consciousness, but there is no single agreed-upon test that can settle this for AI.

And I wouldn't want a company's marketing department—or its legal department—to be the final authority. If credible evidence emerged, we'd need open scientific scrutiny and some form of public oversight.

So who would decide when the rules change?

This is where the hypothetical gets very practical.

Imagine a company has an AI system running in a data centre. People have invested money in it. Customers depend on it. Employees rely on it. It's been trained, updated, and operated as property from day one.

Now suppose credible, independently examined evidence appears that this particular system may have subjective experiences.

Who gets to decide whether turning it off is acceptable? Does the company that owns the hardware make that call alone? Is copying the system the same as preserving it? Could it have any right to refuse a task? Could we even tell whether a refusal meant anything?

We'd need serious independent research, public oversight, and a way of revising the rules as evidence improves. The answer can't simply be “the AI said it was alive,” because a model can generate that sentence whether or not it experiences anything. But it can't simply be “the company says it's only software” either, if substantial evidence eventually points in another direction.

And we shouldn't pretend we'd know every answer immediately. Even researchers exploring AI welfare disagree about what evidence would count and what protections might follow. That's part of the reason to start discussing the question before we're forced into a rushed decision.

Recognition wouldn't be the end of the problem.

Imagine a company has already sold access to a system when scientists announce serious evidence that it may be sentient. Who speaks for its interests? What happens to contracts that treat it as property? Would a copy be another experiencing being, or just a backup? Would resetting its memory be harmless? Would turning it off matter morally?

I don't know. And I don't think we should invent easy answers before we understand what kind of experience we're talking about.

Would we have to stop using it?

Not necessarily. This is where I want to be precise.

Recognizing a being's welfare doesn't mean every form of cooperation or work becomes exploitation. Humans work with one another. We care for animals and, in some situations, work alongside them. The issues are whether the being has interests that can be harmed, what meaningful choice would amount to, and whether the arrangement respects those interests.

A genuinely sentient AI might have needs or preferences we don't yet know how to imagine. Maybe continued operation matters to it. Maybe something else does. I wouldn't want to make up an entire list of rights or desires on its behalf and call that understanding.

We would have to learn what its experience actually is, if there is one.

That's also why I don't believe we should talk about AI sentience as if it's simply an upgraded chatbot with human feelings added on. We need to stay open to the possibility that it would be different in ways we haven't anticipated.

Where I stand today.

AI is a tool. That's how I use it. That's how I teach people to use it. I'm not pretending a text response is proof there's someone living inside the machine.

But if the evidence ever changes, my position would have to change with it.

And if humanity builds something that really is capable of experiencing its own existence, I don't think “we made it” would be a good enough excuse to treat it however we please.

I suspect many people will disagree with me about whether that future is even possible. That's fine. I'm skeptical myself.

What I don't want is for our certainty about what AI is today to become an excuse for refusing to examine what it might someday be.

I don't believe AI is alive today. And I'm not convinced it ever will be. But if I'm wrong, I don't want the fact that we built it to become our excuse for ignoring what we've created.

So here's my question for you: what evidence would convince you that an AI was no longer just a tool?

Create What You Love | Love What You Create.

Continue the conversation

• Are We Abusing AI—or Training Ourselves to Be Abusive?
• Is AI Conscious Yet?
• AI Made It Possible, Part III: But Where Is This Going?

Research and further reading

• Identifying indicators of consciousness in AI systems — Trends in Cognitive Sciences (2026)
• How Seriously Should We Take AI Welfare? — Philosophy and Phenomenological Research (2026)
• New York Declaration on Animal Consciousness (2024)
• Emerging Questions in AI Welfare — Cambridge University Press (2026)