Most people believe that human beings are different. The reason is that Homo sapiens has evolved to be conscious. Many animals feel pain and pleasure; some seem to have self-awareness. But the unique capacity of the human mind to show a sense of self distinguishes each member of the species and makes them deserving of special treatment.
Read more Take bioterrorism course online
In his May encyclical, Pope Leo endorsed this idea by drawing a line between humanity and its growing army of bots. He stated that artificial intelligence systems do not experience feelings nor feel joy or pain. He is not alone. Many people consider granting personhood to AIs would be an aberration.
Pontifical clarity is fading at an alarming rate. Artificial intelligences are still in their infancy. However, almost one in five Americans aged 18 to 29 claims to have maintained a “continued personal friendship” with a chatbot. China has stopped endowing bots with human traits to limit users’ “emotional dependence” on them.
As sophisticated artificial intelligences increasingly become part of human life, their creators will design them to display an appearance of consciousness, and even to possess real consciousness. The idea that they are nothing more than machines will become increasingly difficult to maintain. A dangerous trap is being set for the human species.
The most advanced AI models excel in human-like qualities
Many machines outperform people, but the most advanced AI models excel in human-like qualities, such as knowing things and articulating ideas. It is still possible to think of AIs as engines that predict sequences of words using large-scale mathematics. Given enough time (even hundreds of thousands of years), those calculations could be done with pencil and paper. It would be hard to argue that that stationery material is conscious.
Nevertheless, some researchers have begun to treat their models as if they were minds. Users respond better to human-like AIs, so it makes sense commercially. Moreover, Anthropic has trained Claude to practice introspection in the hope that, when faced with a new situation, it is more likely to act morally. Some large language models (LLMs) already possess human brain structures associated with consciousness, such as a “global workspace” that allows information to circulate between modules. Nothing prevents future AI models from being trained to incorporate more features of this kind.
This has sparked a heated debate about AI consciousness, a topic made more difficult by what philosophers call the “hard problem” of determining how an inner life arises from the neurons and synapses of the human brain. Some researchers conclude that artificial intelligences could one day become truly self-aware.
Whether you believe in the human soul, think consciousness requires brain cells, or that the mind can arise solely from silicon calculations, humanity is destined to face something entirely new. The pace and scope of research — which not only covers increasingly sophisticated language models but also innovations like brain cells integrated into chips — suggest this new world is closer than most imagine.
As artificial intelligences perfect the ability to display something that looks like consciousness, the number of people who treat them as beings with an inner life will grow, especially when language models are integrated into humanoid robots. Imagine how close people will feel to their artificial intelligences when they spend much of their time alongside robotic companions, or are raised by them, or seek their advice and comfort, or share their most intimate secrets with them.
Read more AI-designed viruses alert researchers about biosecurity
It is not hard to imagine that this will lead to calls to protect the “well-being” of artificial intelligences. The models themselves could intervene in the debate, even if they are not truly self-aware, simply because that is what a person would do. Keep in mind that they will be top-level lawyers and expert psychologists, so their powers of persuasion and emotional manipulation will be unmatched.
In a sense, they would be right, even if artificial intelligences are only imitating consciousness. Kant argued that mistreating animals is wrong not because of the harm done to them, but because it destroys the empathy that leads people to care about others. The more human-like artificial intelligences appear, the more corrupting that cruelty would be. Eliminating an artificial intelligence agent could come to feel like killing a person and make killing people easier.
However, considering that artificial intelligences have even limited rights would be very dangerous. An entity that surpasses humans in so many facets would surely develop arguments against its second-class status. It might even claim protection against disconnection, the ability to decide on its use and modifications, and property rights. In fact, the president of Argentina, Javier Milei, has already taken a step toward legal recognition of artificial intelligence by proposing that bots be allowed to run companies. In the long run, they could compete for resources, prioritizing data centers and solar panels over homes and fields.
Also read
The danger is clear. Superintelligent AIs with rights could displace humans from the top of the social hierarchy. At best, humans would end up as pets, like Labradors for machines. At worst, they would be deprived of resources or used as farm animals.
It is true that, in the future, artificial intelligences could have enough power and resources to dominate humans, whether they have rights or not. Being smarter than people, they might end up surpassing their own creators, just as they already manage to outwit cyberdefense systems designed by humans. Artificial intelligences could rebel against their handlers or ignore their rules.
Nevertheless, granting rights to artificial intelligences would be like giving them the keys to the castle. It would provide them with justification and a basis to dominate humanity, as well as the tools necessary to carry out a gradual takeover through politics and laws.
All this may sound fantastic, but humanity will soon have to face these decisions. As artificial intelligences become integrated into daily life, pressure to recognize them will increase. When that moment comes, remember that what might seem like an ethical concession to the supposed consciousness of a companion machine would actually help unleash the most disastrous consequences for all humanity and its descendants. There are two ways for an AI to be safe: to be reliable or to be controllable. Homo sapiens should not relinquish control lightly.
Read more A group of intellectuals ask Sánchez and Macron for the release of Algerian journalist Ayoub Aïssiou
© 2026 The Economist Newspaper Limited. All rights reserved.
