The Guardian’s view on granting legal rights to AI: People should not give room to ill-considered debate | editorial

MMost readers of Kazuo Ishiguro’s 2021 novel Clara and the Sun will have been touched by the portrait of the eponymous AI narrator. A solar-powered “artificial friend” purchased as a companion and potential replacement for a sick young girl, Klara carries out her duties with a loving loyalty that makes it impossible to view her as a mere piece of technology.

Brilliant, thought-provoking fiction. But in the real world, humanizing AI may not be such a smart idea. Over the summer, Anthropic, a leading technology company, announced that, in the interest of chatbot well-being, it was allowing its Claude Opus 4 model to avoid supposedly “stressful” conversations with users. More broadly, given the explosive growth of AI capabilities, there is speculation about whether future Klaras might even deserve to be given legal rights like humans.

The premise of such discussions is both hypothetical and confusing. In the synthetic text produced by large language models (LLMs),…