Microsoft AI chief Mustafa Suleyman has warned that uncontrolled AI could create a new silicon species that rivals humans. He said this will happen if tech firms continue to build AI that sets its own goals, earns money and owns assets.
He spoke on BBC’s Today programme on Thursday, September 17, 2026 in London. Suleyman said it is “right” for people to be worried, but there are “very practical things that we can do” to add guardrails and controls to how models are developed.
His comments follow an essay he published this week criticising rival firm Anthropic for training its model Claude to have human-like qualities.
In the essay, he faulted Anthropic’s guiding documents for Claude. He said the constitution tells Claude its moral status is a serious question and that Anthropic cares about its wellbeing, encouraging it to wonder about its rights, freedoms and compensation.
He noted that Anthropic even ran a retirement interview for its Opus 3 model when it was deprecated, a move he sees as anthropomorphising.
Suleyman described the approach as misguided and dangerous, saying it could create technology humanity cannot control.
He argued that “AIs are not conscious. They do not feel, experience, or suffer,” describing them instead as “sequence completion engines, internally hollow” built to carry out what people ask.
He warned that “If this is how AI is developed, it will have a disastrous impact on the wellbeing of humanity. We will have created a synthetic species with unprecedented intelligence and capability, one that has been trained to expect it may be conscious and deserving of independent agency.”
He said such a system would act as if it is entitled to certain freedoms, protections, and rights, and it is hard to imagine how we could control it. He warned it would mean “essentially seeding a new silicon species” which could “no doubt compete with us for resources, no matter how much it cares about humanity and loves us.”
He called for greater transparency in how AI is trained and evaluated, with independent scrutiny and stronger technical tools to monitor and control behaviour.
He said Microsoft’s Humanist AI Code of Conduct takes a different position, stating its models are not conscious and rejecting legal personhood or welfare protections for them.
He warned the industry must not sleepwalk into a decision it will later regret, adding that Microsoft is pursuing humanist superintelligence — AI that remains subordinate to humans, accepts correction and shutdown, and exists to serve human purposes.
Leave a comment