The creator of the artificial intelligence Claude, Anthropic, announced that starting November 12, it will prohibit “gratuitous and repeated cruelty” toward its AI. This measure aims to sanction insults and malicious behavior directed at Claude AI, a rule that may seem surprising.
According to RFI, this decision marks a profound shift in how laboratories consider the well-being of AI models. It is no longer merely a philosophical question but a security issue. Philosopher and YouTuber Thibaut Giraud sheds light on this evolution, which could reflect a precautionary principle or a risk of anthropomorphism.
This new policy from Anthropic highlights the growing importance given to the relationship between humans and artificial intelligences and raises the question of how these entities should be treated within the framework of their use.
