Microsoft AI chief warns uncontrolled systems could become a rival “silicon species”
Mustafa Suleyman says highly autonomous AI could become difficult to control and compete with humans for resources, calling for transparency and independent scrutiny.
Microsoft AI chief Mustafa Suleyman has warned that artificial intelligence developed without adequate safeguards could eventually become a “silicon species” competing with humans for resources. His comments were reported by BBC World as debate intensifies over the level of autonomy advanced AI systems should receive.
Suleyman said firms should be wary of creating systems capable of setting their own objectives, earning money and owning assets. He described that prospect as a hypothetical risk, arguing such systems could become difficult for people to control regardless of whether they were designed to benefit humanity.
He also criticised Anthropic’s approach to training Claude, saying the company was anthropomorphising the system by presenting it as having human-like desires, values or a sense of self. Suleyman argued that AI systems do not feel or suffer, and should instead remain focused on instructions and goals set by humans. Anthropic’s response was not included in the source material.
Suleyman called for greater transparency about how AI systems are trained and evaluated, alongside independent scrutiny of their behaviour and stronger monitoring and control tools. He said alignment measures should ensure advanced AI remains “subordinate” to humanity.
The warning reflects Microsoft’s own stated approach to advanced AI. The company formed a superintelligence team in October 2025 and has published a draft Humanist AI Code of Conduct describing advanced systems that work for people, stay within limits and remain under human control. Whether AI can develop consciousness, desires or independent motivations remains unresolved.
