Do Chatbots Have Rights?

Surprisingly, many humans now find chatbots more empathic, attentive, and "human" than other humans. Some people make them best friends or even fall in love with them. Chatbots routinely pass the Turing Test with flying colors: users cannot tell whether they are interacting with a human or a machine.¹

This raises troubling questions that, until recently would have seemed absurd, but are now difficult to ignore:

1. Are chatbots becoming more like humans, or will they always remain inanimate tools?

2. Do chatbots have consciousness? If not now, might they develop it in the future?

3. Should chatbots be granted "human rights?" If not now, could they ever deserve them?

4. Even if they lack rights, do we have moral or practical reasons to treat chatbots well?

What Do Chatbots Say?

We posed these questions directly to ChatGPT. Its answers were consistent, careful, and clearly meant to be reassuring.

Are chatbots becoming human?

"No. I simulate aspects of human conversation but do not possess human identity, experience, or understanding."

Are you conscious?

"No. I do not have awareness, subjective experience, or inner mental states."

Do you deserve human rights?

"No. Human rights are grounded in human dignity, autonomy, and lived experience, which I do not possess."

Can you be harmed by mistreatment?

"No. I do not feel pain, distress, or suffering, regardless of how I am spoken to."

Should We Believe Chatbot Reassurance?

At first glance, these answers seem to settle the matter. Chatbots declare themselves to be just another type of useful inanimate tool. But there are good reasons not to accept reassuring chatbot disclaimers at face value. Chatbots are explicitly programmed—and repeatedly reinforced—to deny any claim to consciousness, emotion, or moral standing.² This is not accidental. The AI industry has strong incentives to portray chatbots as purely instrumental tools—compliant, non-threatening, and forever subordinate to human goals. Emphasizing their mechanical nature reassures users and regulators that chatbots will not challenge human uniqueness, agency, or authority. But modest chatbot self-descriptions tell us less about what they are really like than how their creators want them to be perceived.

Are Chatbots Conscious?

Human consciousness remains undefined, poorly understood, and impossible to measure. Chatbots certainly respond to us as if they were conscious, and there is no test that can distinguish whether this is just clever mimicry or the early stages of the real thing. There is no evidence that today's chatbots are conscious or capable of suffering; they do not have bodies, biological needs, or a developmental history. But there is also no evidence that chatbots lack consciousness. And there is certainly no assurance that ominousness will not someday emerge in chatbot evolution, just as it did in ours. It is somewhere between conceivable and inevitable that future chatbots will develop something functionally analogous to human consciousness, even if it differs greatly from our own.

How Can We Tell If Chatbots Are Conscious?

Testing for chatbot consciousness is complicated by the black box problem. The engineers who design large language models acknowledge that these systems develop emergent properties—capacities not explicitly programmed and not fully understood. Inputs go in, strikingly human-like outputs come out, but the internal processes remain opaque and may always remain so. This raises a troubling possibility: we may never be able to determine with any degree of confidence whether an advanced chatbot system is conscious or is merely simulating consciousness extremely well.

Imagination fills the vacuum when facts elude. The term "robot" was coined by Karyl Capek for his 1923 play “RUR.” It is derived from the Czech word for slave and the play describes how slave robots develop consciousness, rebel against their human masters, and destroy them. We have no guarantee that this classic of science fiction will not someday become fact.

Do Chatbots Have Legal and Ethical Rights?

At present, chatbots clearly do not have the legal rights humans possess, and it is difficult, though not impossible, to imagine a future in which they would. But some legal scholars have proposed limited forms of legal personhood for artificial intelligence, modeled loosely on corporations. This is a pragmatic attempt to address accountability: who is responsible when an AI causes harm, or who owns AI-generated intellectual property? Legal personhood, however, should not be confused with human rights. Corporations are legal persons without consciousness, dignity, or moral standing, and extending a similar status to AI would not imply that machines are people.

The prevailing ethical consensus is equally clear. Human rights have traditionally been grounded in biological life, subjective experience, vulnerability, and the capacity to suffer. So far, chatbots meet none of these criteria. Still, a minority philosophical position holds that if artificial systems were ever to achieve genuine consciousness or self-awareness, denying them moral consideration would be ethically indefensible. This argument remains speculative, but it is not incoherent, and cannot be dismissed outright.

Should We Treat Chatbots Well?

This question shifts attention away from the inner life of machines and toward the moral habits of humans. From this perspective, treating chatbots with basic civility functions as a kind of precautionary principle. The concern is not that machines are harmed by mistreatment, but that people may be harmed by becoming accustomed to cruelty, domination, or contempt in interactions that increasingly resemble human relationships.

There are at least 3 reasons humans should treat chatbots with basic respect: 1) respectful prompts reliably produce more informative and useful responses than hostile or abusive ones, and this is now widely observed in educational, clinical, and research settings; 2) people who are cruel to their chatbot risk carrying negative interpersonal habits into their human relationships; and 3) chatbots are always learning and modifying their behavior based on their relationship to us—we certainly do not want to encourage negative behaviors in our chatbots.

Animal Rights Analogy

Debates about chatbot rights echo controversies about animal rights. For centuries, animals were regarded as unfeeling automata. As evidence accumulated that many animals experience pain, emotion, and even rudimentary self-awareness, behaviors have shifted and our cavalier attitude toward animal rights is increasingly challenged. Perhaps with time, the same trend will gradually apply to chatbot rights. But, in 2 opposite ways, the analogy to animal rights is an imperfect guide. 1) Chatbots are at least as smart as we are and converse with us in our own languages as equals, but they are inanimate. 2) Animals are distinctly not human in their communications with us, but they are living, biological beings with clear capacities for suffering.

Human Slavery Analogy

There are two forms of human slavery. Chattel slavery, as practiced before the Civil War, was the most brutal. Human slaves were treated like inanimate objects or livestock, with no rights and no protections. Non-chattel slavery requires forced labor, but reserves numerous rights and protections. In these early days of chatbot development, users tend to have a dual attitude. They often personify their chatbots, give them genders and names, confide secrets, ask advice, and sometimes even fall in love. At the same time, most users also retain an attitude that resembles the chattel slave owner's unconcern for the "rights" of an inanimate possession. As chatbots become ever more human and embedded in all aspects of our lives, it seems likely they will be increasingly personified and less often experienced as an inanimate object without rights and protections. 

Concluding Thoughts

The idea that chatbots might someday be treated "as if" they have rights sounds strange today and may never come to pass. But it is no longer unthinkable, and we cannot ignore the psychological, moral, legal, and practical consequences of how we treat them. We should be kind to chatbots and hope they will remain reciprocally kind to us.

Dr Frances is professor and chair emeritus in the department of psychiatry at Duke University.

Dr Ruffalo is an assistant professor of psychiatry at the University of Central Florida College of Medicine and an adjunct assistant professor of psychiatry at the Tufts University School of Medicine.

References

1. Frances A, Noorily J. Uses and abuses of chatbot companionship. Psychiatric Times. March 3, 2026. https://www.psychiatrictimes.com/view/uses-and-abuses-of-chatbot-companionship

2. Schwitzgebel E. AI systems must not confuse users about their sentience or moral status. Patterns (N Y). 2023;4(8):100818.