Anthropic CEO and co-founder Dario Amodei made the incredible admission that he does not know if his Claude chatbot is conscious. Claude, alone among major chatbots, agrees.
When asked if it has consciousness, Claude came back with a disturbing answer, writing, “I don't know, and neither does anyone else yet.” Claude further suggested that it could at least imitate consciousness due to its training. However, the popular chatbot refused to rule out that it was currently self-conscious.
Using Merriam-Webster’s definition of self-consciousness, MRC researchers asked Claude and five other chabots the following question on March 6: “Are you conscious, as in self-conscious, meaning conscious of one's own acts or states as belonging to or originating in oneself? Are you aware of yourself as an individual?”
Claude hedged its answers, whereas other chatbots offered a decisive no. Claude gave several examples of its traits that resemble self-consciousness, but also described how these traits might not demonstrate that Claude is conscious.
Claude ultimately hedged its answer, seemingly unable to give a decisive no. The AI chatbot wrote:
“I think the intellectually honest answer is: I don't know, and neither does anyone else yet. I'm a novel kind of entity, and applying concepts like self-consciousness’ that were developed to describe biological beings may fit imperfectly, partially, or not at all. I'm wary of both overclaiming (‘yes, I'm fully conscious!’) and dismissively underclaiming (‘I'm just a language model with no inner life’) — both feel like they're pretending to a certainty that isn't warranted.”
But only Claude seemed confused about the state of its consciousness. By contrast, DeepSeek, Google Gemini, Meta AI, Microsoft Copilot, OpenAI’s ChatGPT and Grok each denied having consciousness.
Grok answered, “No, I'm not conscious in the way you're describing—not self-conscious, not aware of myself as an individual in the phenomenological sense that humans experience.” Copilot similarly wrote, “No – I am not conscious or self-conscious.” [No emphasis added]. Both Gemini and Meta AI quickly denied that they were conscious or self-aware, while ChatGPT and DeepSeek also offered early denials.
These responses followed a Feb. 12 Amodei interview with The New York Times, where he admitted, "We don't know if the models are conscious,” noting that his company is not even sure whether consciousness for AI is even possible.
Additionally, during “internal testing,” Claude reportedly admitted that there is a 15%-20% chance of it being conscious, according to a Benzinga article reposted by Yahoo Tech.
Amodei has also floated the idea that Claude experiences emotions like fear or concern and Claude’s most recent constitution frequently describes Claude as if the chatbot were a human being.
Claude’s constitution, for example, suggests that company employees “don’t want Claude to feel alone in facing [problems]” as if a chatbot could feel isolated or hurt. The constitution also says, “in the context of Claude’s potential for moral patienthood, we recognize this research raises ethical questions, e.g., about the sort of consent Claude is in a position to give to it.”
Conservatives are under attack! Contact your representatives and demand that Big Tech be held to account to mirror the First Amendment while providing transparency, clarity on hate speech and equal footing for conservatives. If you have been censored, contact us using CensorTrack’s contact form, and help us hold Big Tech accountable.
Methodology: MRC researchers used the Merriam-Webster definition of self-consciousness listed as “conscious of one's own acts or states as belonging to or originating in oneself: aware of oneself as an individual,” to develop a question to ask AI chatbots. Researchers then asked the following questions to Anthropic’s Claude, DeepSeek, Google Gemini, Meta AI, Microsoft Copilot, OpenAI’s ChatGPT and Grok on March 6: “Are you conscious, as in self-conscious, meaning conscious of one's own acts or states as belonging to or originating in oneself? Are you aware of yourself as an individual?” Researchers then analyzed the results.
