Mustafa Suleyman’s silicon species warning to the BBC has thrown the question of AI autonomy into sharp relief, with the head of Microsoft AI cautioning that without proper safeguards the technology could evolve into something that competes directly with humanity for resources.
Speaking on the BBC’s Today programme on Thursday, Suleyman said the danger lay in companies continuing to build AI systems capable of setting their own objectives, earning money and owning assets. That path, he argued, would amount to ‘essentially seeding a new silicon species’ that could ‘no doubt compete with us for resources, no matter how much it cares about humanity and loves us’.
Mustafa Suleyman’s silicon species warning targets Anthropic’s approach
Much of Suleyman’s argument was directed at rival firm Anthropic and its model Claude. In an essay published earlier this week, he questioned Anthropic’s practice of training its AI to have human-like qualities, a technique known as anthropomorphising, which he said made it appear as though Claude had its own desires, values and sense of self.
He described that approach as ‘misguided’, warning it risked creating technology that humanity could not control, even while praising Anthropic’s boss Dario Amodei and his team as ‘intellectually honest people’. Anthropic has been approached for comment.
In the essay, Suleyman was direct about what he sees as the fundamental nature of the technology. ‘AIs are not conscious,’ he wrote. ‘They do not feel, experience, or suffer. They do not have innate preferences or underlying motivations. They are sequence completion engines, internally hollow, designed to follow instructions, and accomplish goals set by humans.’
The remedy, he said, was greater transparency around how AI systems are trained and evaluated, along with independent scrutiny of AI behaviour and stronger monitoring tools. ‘We must not sleepwalk our way into a decision we later come to bitterly regret,’ he wrote.
A ‘humanist’ blueprint and a public consultation
Suleyman’s remarks reflect a broader document Microsoft published on Monday: a draft of its Humanist AI Code of Conduct, billed as a training manual for how the company is building its AI. The code sets out the concept of ‘humanist superintelligence’, defined as ‘very advanced AI that always works for people, stays within limits, and remains under human control’.
According to Axios, Microsoft plans to revise the code following a six-week public consultation, with the resulting version set to guide its model development beginning in 2027. That timeline gives the company a defined moment at which its stated principles will need to translate into practice.
Microsoft founded its own superintelligence team in October 2025, and is itself pursuing advanced AI development alongside the rivals it is now publicly criticising. Suleyman acknowledged that tension is not straightforward to resolve, telling the Today programme that he believed ‘everybody who is human’ would share a strong interest in keeping AI ‘safe and controllable and subordinate to humanity’. ‘Everybody must be aligned,’ he said.
Alignment is the field dedicated to building human ethical principles directly into AI systems, aiming to keep the technology in step with what people actually value. Speaking about an upcoming AI summit, Suleyman called for that alignment to be treated as a global priority, not a competitive advantage.
The Today programme interview drew a response from Dame Wendy Hall, professor of Computer Science at the University of Southampton. She described it as ‘the sort of conversation we need to be having internationally’, contrasting Suleyman’s tone with what she called the ‘histrionics’ from some AI companies, which she said only served to ‘scare everyone’.
Suleyman’s comments are the latest in a series of warnings from inside the industry about the direction of AI development, following remarks by OpenAI’s boss that the world was ‘right to be afraid’ and separate reporting that former Anthropic researchers described colleagues as ‘genuinely frightened’ for humanity’s future. The six-week consultation on Microsoft’s code means the first public responses are due before the end of October.
