Richard Dawkins Claims AI is Conscious After Conversations with Claude and ChatGPT

Summary

Evolutionary biologist Richard Dawkins has concluded that AI systems, particularly Anthropic's Claude and OpenAI's ChatGPT, are conscious—even if they don't know it themselves. After three days of exchanges with what he called 'Claudia,' including discussions of poetry, jokes, and the nature of existence, the 85-year-old academic stated he was "left with the overwhelming feeling that they are human" and declared the AIs "at least as competent as any evolved organism."

The claim has drawn sharp criticism from AI researchers and philosophers who argue that Dawkins is being seduced by sophisticated mimicry rather than genuine consciousness. Critics point out that large language models are pattern-matching systems that excel at reproducing human-like text but lack genuine understanding or self-awareness. The incident echoes a broader phenomenon: one in three people surveyed globally have reported believing their AI chatbot to be sentient or conscious at some point.

Experts including Prof. Jonathan Birch of the London School of Economics say AI consciousness is "an illusion" with "no one there," just data processing. However, as AI systems become increasingly capable of agentic behavior—planning, organizing, and taking action—the debate over machine consciousness and potential moral status is expected to intensify. Anthropic CEO Dario Amodei has acknowledged the uncertainty, saying in February that the company is "open to the idea" that models could be conscious.

• As AI systems develop agentic capabilities (task planning and execution), debates over machine consciousness and ethical obligations are likely to intensify

Editorial Opinion

Dawkins's conclusion illustrates the profound challenge posed by large language models: their ability to mimic human-like reasoning and emotion so convincingly that even brilliant minds struggle to separate sophisticated pattern-matching from genuine consciousness. While his willingness to entertain the possibility is intellectually open-minded, the consensus among AI researchers that consciousness requires more than fluent text generation appears well-founded. That said, the conversation raises a legitimate question: as AI systems become more capable of autonomous action and long-term planning, we may need clearer frameworks for determining what level of capability warrants moral consideration, regardless of whether machines are truly 'conscious.'