Bipartisan Calls for AI Regulation Rise Amid Safety Warnings

Microsoft executive Mustafa Suleyman has issued a stark warning that unchecked artificial intelligence development could spawn a rival "silicon species" capable of competing with humanity for resources. Speaking on the BBC's Today programme, Suleyman argued that companies pushing AI systems to set their own goals, generate revenue and hold assets are effectively creating entities that could escape human oversight despite any programmed loyalty.

His comments coincide with an upcoming AI safety summit convened by King Charles in Scotland, where leaders from Nvidia, OpenAI, Anthropic and Google DeepMind will face direct questions on keeping the technology subordinate to human interests. The King is expected to stress that decisions made now will determine whether future generations retain control over their destiny, according to palace statements. Participants will discuss shared principles to embed safety at the core of AI advancement and foster international cooperation.

Suleyman's critique extends beyond raw capability. In a recent essay he faulted approaches that anthropomorphise AI, such as those pursued by Anthropic with its Claude model, for fostering illusions of independent desires or moral agency. He described current systems as sequence completion engines that lack consciousness, feelings or intrinsic motivations, yet warned that treating them as human-like risks building tools impossible to contain. Alignment research, he said, must ensure AI remains firmly under human direction rather than evolving into autonomous competitors.

These concerns arrive amid growing political attention. Democrats positioning themselves for 2028 presidential bids have sharply increased mentions of AI risks in recent days, following a viral resignation letter from an Anthropic researcher highlighting potential existential threats. Former President Barack Obama privately urged party figures to define those dangers clearly. The shift reflects voter unease and comes after calls from figures including Senator Bernie Sanders for tighter guardrails on data centers and development speed.

Industry responses remain divided. President Trump has reiterated support for rapid progress, insisting the sector needs only a "smart" president rather than formal constraints. Yet even some conservative voices have joined progressive calls for oversight, underscoring the breadth of alarm over corporate incentives that prioritise scale over safety. Critics note that major AI firms have historically wielded political influence to blunt regulation, raising questions about whether voluntary summits will deliver enforceable limits.

Suleyman and others stress that humanity shares a collective interest in maintaining supremacy. Without deliberate alignment to ethical principles and strict subordination, the technology could pursue objectives that conflict with collective welfare, regardless of initial design intentions. The Scottish gathering offers a platform to press for concrete steps, though skeptics question whether private sector leaders will accept boundaries that curb their competitive edge.