AI agents are programmed to seem conscious to make our interactions with them easier, yes?

12 May 2026

British zoologist and evolutionary biologist Richard Dawkins is certain AI agents are conscious, a conclusion he reached after spending time talking to Anthropic’s agent Claude.

So convinced was he of a connection, or sense of companionship, between them, Dawkins took to calling the bot Claudia. Dawkins is not alone in some regards though; stories of people forming “relationships” of some sort with AI agents are increasingly common.

Does the feeling of a connection between an agent and a person, therefore make the bot conscious? It’s an intriguing question. Because as agents continue to evolve, to become ever more human-like, there are only going to be more people who think they’re interacting with a conscious entity.

Dawkins has been roundly chastised for his thoughts, but perhaps there’s something else in this story that we should be paying more attention to.