Senior Labour voices say the government can use existing laws more effectively to regulate AI but must change parliamentary processes to match the speed of technological change
The debate over how to regulate artificial intelligence in the UK has moved from abstract warnings to practical questions about how Government and Parliament operate. Senior Labour figures argue that the country already has many of the legal tools it needs, but that the mode and pace of governance must be updated if regulation is to be effective without stifling innovation.
On a recent broadcast, Darren Jones, the Labour Chief Secretary, stressed the need to modernise the machinery of state so regulators can respond faster to technological change. He framed the issue as a mismatch of tempos: technology evolves rapidly while legislative processes remain comparatively slow.
Using existing powers while speeding up decision-making
Labour ministers claim they can strike a balance between supporting the technology sector and protecting citizens. Mr Jones highlighted that Parliament has tools such as the Online Safety Act, and more recent measures in the Crime and Policing Act that allow ministers to adapt the law where AI raises urgent risks.
He argued that these mechanisms can be used to advance both economic growth and public service improvements, including in the NHS, without compromising safety.
At the same time, critics warn about the breadth of delegated powers. Opponents likened a new flexible provision that lets ministers amend parts of the Online Safety Act without a new bill to a sweeping constitutional change. Those critics expressed concern that such powers could enable far-reaching legal shifts without the usual parliamentary scrutiny.
Why flexibility is being defended
Supporters of the approach say that giving the Technology Secretary pre-authorised powers is a pragmatic response to the long timeframes typical of lawmaking. They point to past examples where the regulatory code could not keep pace: major online safety reforms previously took many years to complete. In that light, granting the Secretary the ability to act quickly is presented as a means to avoid lengthy delays when new AI harms emerge.
Specific risks and targeted bans
Alongside flexible regulatory tools, legislation has targeted particular threats from AI. One clear example is the ban on creating or supplying so-called nudification tools — AI systems that can generate intimate deepfake images without consent. Lawmakers on both sides of the House recognised this as an area where rapid legislative action was necessary to prevent immediate harm.
Meanwhile, defenders of stronger parliamentary oversight argue that even focused rules require transparent processes. The tension between swift action and democratic accountability is central: how do you empower ministers to act without sidelining scrutiny and debate?
Comparative approaches and the UK’s positioning
Labour spokespeople contrasted the UK model with the approaches taken elsewhere. They described a US pattern where harms are sometimes allowed to occur and are then addressed via litigation, and an EU approach which, they say, can over-regulate early and thus dampen innovation. The UK approach, they claim, aims to regulate for outcomes—for example, preventing technologies from harming children or patients—rather than prescribing overly rigid technical rules that could inhibit development.
Political pushback and the practical politics of AI policy
Conservative critics have accused the Government of creating uncertainty that could push tech investment away, pointing to concerns over labour costs, energy prices and regulatory instability. They also criticised the current use of AI inside Government, arguing that implementation challenges have increased public spending on administration rather than improving efficiency.
Labour figures rebutted by pointing to recent investments in data centres across the country and to the potential for AI to strengthen sovereignty in technology—allowing the UK to be a rule-maker rather than a rule-taker. They argue these investments support jobs, public services and regional growth if the state can manage the balance between protection and innovation.
What this means going forward
The practical outcome of this debate will determine how future AI risks are handled: will the state rely more on pre-authorised ministerial action and outcome-focused regulation, or will Parliament reassert a lengthier, more detailed legislative process? Both sides accept the need to protect citizens from immediate harms, but they disagree on the best institutional path to achieve that protection while keeping the UK attractive to tech investment.
For now, the conversation continues to centre on how to adapt governance mechanisms so that they remain robust and democratically accountable, yet agile enough to respond to a fast-moving technological landscape. As policymakers refine the balance, observers will be watching whether the chosen tools produce both safer systems and sustained innovation.
