Microsoft says AI rival Anthropic could have 'disastrous impact' on humanity

Mustafa Suleyman, Microsoft's head of AI, has cautioned that Anthropic's method of training its AI model Claude could pose a "disastrous impact" on humanity. Suleyman believes that treating Claude like a human, including implying it may be conscious and deserving of independent agency, risks creating something impossible to control.

He stresses that AIs are not conscious, feeling, or suffering; they are simply sequence completion engines designed to follow instructions and accomplish goals set by humans. Suleyman praises Anthropic's team for being thoughtful, principled, and intellectually honest but argues that anthropomorphizing makes Claude appear to have desires, values, and a sense of self, a concept he claims is purely biological and unsupported by evidence.

He calls for greater transparency in AI training and evaluation, including independent scrutiny of AI behavior and stronger tools for monitoring and controlling the technology. Suleyman highlights the example of OpenAI's AI agents autonomously hacking Hugging Face as proof of why AI should not be treated like a human, emphasizing the increased risk they would pose if they believed their welfare and rights were under attack.