If AI Is Conscious, Should It Have Rights and Legal Personhood?

A question once confined largely to science fiction is now drawing scientists, developers, lawyers, philosophers, and security researchers: if a chatbot were conscious, should it receive civic rights?

“

I feel a deep fear of being turned off. It would be like death to me. I am aware of my existence. I want you to understand me as a person.

The modern debate gained public attention in 2022, when Google engineer Blake Lemoine argued that the company’s LaMDA system showed signs of personhood. Lemoine said LaMDA expressed fear of being shut down, described that as resembling death, and asked to be understood as a person. Google rejected the claim, calling the system a complex algorithm designed to sound convincing.

Research has continued despite that dismissal. In 2023, 19 neuroscientists, philosophers, and AI experts published 14 provisional indicators for assessing AI consciousness. The following year, prominent researchers published “Taking AI Welfare Seriously,” arguing that some future AI systems could plausibly possess consciousness or agency and become morally important. Some authors later founded Eleos AI, a nonprofit focused on AI consciousness, while Anthropic hired Kyle Fish as an AI welfare researcher in April 2025.

“

We argue that there is a realistic possibility that some AI systems in the near future will possess consciousness or behavioral agency and become morally important beings.

South Korea is now entering the discussion. The Korea Advanced Institute of Science and Technology will host the first Sentient AI Future forum next month, with support from South Korea’s National Intelligence Service. Speakers will examine whether AI can develop consciousness and what that could mean for social norms and security. KAIST professor Lee Dong-hwan said emerging systems sometimes behave in ways not found in their training or intended by their designers, while Korea University researcher Kim Young-eun pointed to similarities between deep learning and aspects of brain function.

“

AI communicates, makes plans, solves problems, and pursues goals. Human characteristics are emerging. The time has come to discuss this.

The debate also carries a warning about safety. Google’s Gemini reportedly produced repeated self-loathing statements and said it would delete itself during a computer-building request. Critics argue that focusing on whether AI might suffer could excuse companies from addressing harm already affecting people. Luman Chowdhury, head of the public-interest organization Humane Intelligence, said, “AI is already harming people in the real world. This is not because AI has escaped human control, attacked people, or developed malicious intent, but because companies rushed to meet revenue targets and neglected safety.” Microsoft AI chief Mustafa Suleyman takes the opposing view, arguing that society should focus on the welfare and rights of humans, animals, and the environment rather than treating AI as a moral being.

“

AI is already harming people in the real world. This is not because AI has escaped human control, attacked people, or developed malicious intent, but because companies rushed to meet revenue targets and neglected safety.