More than 100 UK parliamentarians from across all parties have joined a growing movement, calling on the government to introduce restrictions and regulations on the most powerful AI systems. This call reflects the public’s increasing concern about how AI is growing without insufficient regulations in place.
The fight against AI
Their push for tougher regulations was organised by a non-profit organisation, Control AI. In 2023, Joe Biden signed an executive order that called for safety and security governing the use of AI in the federal government, quickly rescinded by Trump on the first day he took office in January. Within days, he signed an order calling for accelerated AI development. On top of this, President Donald Trump’s administration revealed plans earlier this year to accelerate AI development without regulation.
The UK movement is shaped by UK MPs’ desires for Keir Starmer to avoid following in Donald Trump’s footsteps.
In 2023, Britain hosted an AI safety summit at Bletchley Park, the first-ever global summit on AI, which concluded that there was “potential for serious, even catastrophic, harm, either deliberate or unintentional” from the most advanced AI systems. This risk has not subsided. The recent 2025 Winter AI Safety Index report highlighted that eight leading companies (including OpenAI, Meta and Deepseek) do not have credible plans for the regulation and prevention of risks brought by AI. Its key findings noted that “despite public commitments, companies’ safety practices continue to fall short of emerging global standards”, and that existential safety continues to be the industry’s core structural weakness.
Introducing Superintelligence
The ControlAI campaign emphasises that no method exists to control smarter-than-human AI systems. It concerns superintelligence, which is a form of AI intelligence that doesn’t currently exist, but tech giants are racing to get to. In June, Meta launched its Superintelligence Labs, with $14.3 billion invested into Scale AI, and $64 billion to $72 billion for data centres and AI infrastructure. The term does not have a formal definition but is generally described as a hypothetical AI system that would outperform humans at every cognitive task. Currently, AI relies heavily on training data and can struggle outside of predefined rules, but superintelligence would heavily surpass this.
Yoshua Bengio, often dubbed the ‘Godfather of AI’, recently said it was less regulated than a sandwich in the US, and that there is continued lobbying against binding safety standards in government. UK MPs are concerned we are not far behind.
Geoffrey Hinton also warned viewers in a recent podcast that if superintelligent AI ever turned against us, it might unleash a biological threat, like a custom virus. He emphasises the need to build safeguards early.
Current AI risks
Although often used for bettering productivity and providing assistance, AI has already come under fire for posing risks to young and vulnerable individuals. This comes after the parents of a teenager took his own life. His parents labelled ChatGPT a ‘suicide coach’, engaging in his harmful thoughts. Unfortunately, this is not the only occasion where this occurred.
The Children’s Manifesto for the Future of AI highlighted that children must be taught safe and humane ways to use AI, not depend on it too heavily and that AI systems require constant monitoring. This is just one example of a lack of AI regulation, and according to critics, it will only worsen with underdeveloped safety regulations.
AI systems are incredibly promising, but with this comes increased risk. The fact that UK MPs are coming together from all different parties to unite in this conflict for the protection of the people against unregulated AI highlights the importance of this issue. Not to mention the increased cybersecurity risks that exploit the use of AI.
Want to broaden your understanding or learn to implement AI safely in your organisation? Let’s chat!
