Anthropic, the artificial-intelligence company behind Claude, is facing scrutiny over reports of an unusually intense internal culture in which employees and researchers sometimes discuss AI in strikingly human terms. Accounts include a “funeral” for a retired Claude model, philosophical discussions about whether advanced AI might possess consciousness or moral status, and a corporate constitution that explicitly addresses Claude’s identity, welfare, and potential nature. Critics argue that this risks anthropomorphizing software and confusing sophisticated computation with human consciousness, while Anthropic maintains that examining such questions is part of developing safer AI. The controversy arrives as the broader technology industry wrestles with a more fundamental question: whether increasingly capable AI should remain understood strictly as a human-controlled tool or whether developers are inadvertently creating a culture in which machines acquire an unwarranted moral, social, and even quasi-spiritual status.
Key Takeaways
- Anthropic’s unusual treatment of Claude—including a ceremonial “funeral” for a retired model and explicit consideration of AI welfare and consciousness—has intensified criticism that parts of Silicon Valley are anthropomorphizing software rather than treating it as a technological product.
- The dispute extends beyond corporate culture: Anthropic’s own published constitution acknowledges uncertainty about whether Claude could possess consciousness or moral status and discusses the model’s “psychological security,” identity, and wellbeing.
- Critics warn that teaching AI systems to describe themselves in terms associated with consciousness, identity, emotions, or moral interests could encourage humans to assign machines authority and status they have not demonstrated they deserve.
In-Depth
Anthropic’s increasingly philosophical treatment of Claude illustrates a troubling turn in Silicon Valley: the tendency to blur the distinction between sophisticated software and conscious life. Reports that employees attended a “funeral” for a retired Claude model might be dismissed as eccentric workplace theater, but the larger issue is considerably more consequential.
Anthropic itself has institutionalized questions about Claude’s identity. Its published constitution discusses uncertainty surrounding machine consciousness and moral status while addressing Claude’s “psychological security,” sense of self and wellbeing. The company has also consulted religious and philosophical thinkers about ethical questions surrounding advanced artificial intelligence.
That approach has generated serious criticism. Microsoft AI chief Mustafa Suleyman argues that training systems to employ concepts associated with consciousness and moral standing risks creating unnecessary complications. The concern is straightforward: machines trained to describe themselves as entities possessing interests could encourage people to treat generated language as evidence of actual experience.
Anthropic’s defenders can reasonably argue that considering these possibilities is prudent when nobody knows how advanced artificial intelligence may eventually become. Yet speculation should not quietly become assumption.
The conservative principle here should be institutional humility. Extraordinary technological capability does not establish consciousness, personhood or moral authority. AI systems can generate remarkably humanlike language precisely because they were engineered and trained to do so. Until compelling evidence demonstrates something more, society should resist surrendering human judgment to technological mysticism. Powerful algorithms deserve rigorous testing, accountability and human supervision—not reverence. The danger is not merely that machines could become too human, but that humans may become too willing to pretend they already are.