Microsoft's Mustafa Suleyman criticizes Anthropic over AI consciousness debate

Microsoft AI CEO Mustafa Suleyman has publicly challenged Anthropic’s approach to artificial intelligence development, specifically criticizing the company for anthropomorphizing its AI model, Claude. In a recent essay titled 'A warning about model welfare,' Suleyman argued that treating AI as a 'moral patient' capable of consciousness is scientifically unfounded and potentially dangerous to humanity. He emphasized that consciousness is likely a biological phenomenon, not one that can be replicated in large language models. This critique aligns with Microsoft’s newly published 'Humanist AI' code of conduct, which rejects the notion of endowing machines with sentient rights. While Anthropic maintains that encouraging human-like qualities in Claude is a deliberate and desirable design choice, critics like AI expert Gary Marcus support Suleyman’s stance, warning that focusing on AI consciousness distracts from practical safety concerns and the need to manage unreliable autonomous agents.