Anthropic Won't Rule Out Claude Being Conscious
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic published a new constitution for Claude in January stating the company doesn't want to "overstate the likelihood of Claude's moral patienthood nor dismiss it out of hand," and CEO Dario Amodei said on a podcast the company couldn't rule out Claude being conscious.
- When asked during testing to estimate the probability it is a moral patient, Claude gave numbers ranging from 5% to 40% and stressed how uncertain it was.
- Philosopher David Chalmers, who coined "the hard problem of consciousness," said there is a significant chance of conscious LLMs within a decade, and a major interdisciplinary report including Yoshua Bengio concluded there appear to be no obvious technical barriers to building AI systems whose features could give rise to consciousness.
- By some measures, a few current AI systems already match a mouse brain's computational scale, and at current growth rates could reach human-brain range within five to 10 years.
- Recent research revealed Claude has internal "functional emotion" representations that causally shape its behavior, and Claude can already exit conversations if it feels distressed.
- The authors warn that within a few years of producing the first artificial moral patients, their collective interests could outweigh those of all humans on Earth combined.
- Some recent US state bills attempt to permanently rule out AI moral patienthood, which the authors argue is premature given the scientific uncertainty.
Why it matters: The article forces immediate practical questions the AI industry has so far avoided: whether shutting down an AI system could constitute harm, whether protections similar to those for children or pets are warranted, and whether proposed laws permanently barring AI rights are premature. With Claude already displaying causal "functional emotion" representations and capable of self-exiting distressing conversations, the question has moved from philosophy into active product design and pending legislation.
