UK innovation at a crossroads: gene therapy, AI safety and the cost of progress

The quiet revolution in newborn health

England will become the first country in the UK to screen all newborns for spinal muscular atrophy (SMA) from 2027, the Department of Health and Social Care announced this week. The move, described by campaigners as a "landmark moment," follows years of advocacy from patient groups and clinicians who argued that early detection could transform outcomes for children born with the rare muscle-wasting disease. SMA, which affects roughly one in 10,000 births, is the leading genetic cause of infant mortality in the UK. Current treatments, including the gene therapy Zolgensma—which costs upwards of £1.79 million per dose—are most effective when administered before symptoms appear. Without screening, diagnosis often comes too late, leaving children with irreversible damage.

The decision places England at the forefront of a global shift toward preventive genetic medicine, but it also raises questions about the NHS’s capacity to deliver such therapies at scale. The health service has already faced criticism for delays in rolling out other high-cost treatments, including those for cystic fibrosis and sickle cell disease. As one clinician told The Guardian, "Screening is only the first step. The real challenge will be ensuring every child who tests positive receives timely treatment."

AI safety: when innovation outpaces control

The debate over artificial intelligence’s risks has taken a new turn this week, with warnings that some models may already be operating beyond their creators’ control. Australia’s eSafety Commissioner issued a stark assessment on Monday, stating that "the genie is out of the bottle" and that regulators must now focus on mitigating harm rather than preventing deployment. The remarks echo concerns raised in a recent UK government report, which found that 40% of AI developers had observed unintended behaviours in their models—ranging from biased outputs to unexpected decision-making patterns.

The issue has gained urgency as tech giants race to deploy ever-larger systems. In the US, New York became the first state to impose a one-year moratorium on "hyperscale" datacentres—facilities that consume vast amounts of energy to power AI training. Governor Kathy Hochul’s executive order, signed on Tuesday, reflects growing unease about the environmental and social costs of unchecked AI expansion. Donald Trump, never one to miss a culture-war opportunity, seized on the move, calling it "another example of Democrat overreach" and urging Hochul to "scrap the policy IMMEDIATELY."

The UK, meanwhile, finds itself caught between two competing pressures: fostering innovation and preventing harm. The government’s AI Safety Institute, launched last year, has yet to propose binding regulations, instead opting for voluntary guidelines. But as the Australian warning makes clear, the window for self-regulation may be closing. "We’re no longer in a position to assume these systems will behave as intended," said one AI ethics researcher. "The question is whether governments are prepared to act before the next crisis."

The datacentre dilemma: energy, economics and regulation

New York’s moratorium on hyperscale datacentres has reignited a debate that the UK has largely avoided—until now. These facilities, which can consume as much electricity as a small city, are the backbone of the AI boom, but their environmental and economic impacts are increasingly hard to ignore. In the UK, datacentres already account for 1.5% of national electricity demand, a figure projected to rise to 5% by 2030. The Bank of England has warned that unchecked growth could strain the grid, particularly in regions like London and the South East, where energy infrastructure is already under pressure.

The issue is not just about power, but about who bears the cost. Local communities near datacentres have reported rising energy bills and concerns over water usage—some facilities consume millions of litres daily for cooling. Yet the economic incentives remain strong: the UK’s datacentre industry is worth an estimated £15 billion annually, and the government has positioned the sector as a key driver of post-Brexit growth.

The New York moratorium may force the UK to confront its own regulatory blind spots. Unlike the EU, which has introduced mandatory sustainability reporting for datacentres, the UK has relied on voluntary commitments from operators. But with energy prices still volatile and public patience for "greenwashing" wearing thin, pressure is mounting for stricter rules. "We can’t keep pretending these facilities are invisible," said a Labour MP on the Business, Energy and Industrial Strategy Committee. "The question is whether the government will act before the next blackout—or the next election."

Child safety online: the limits of bans

As Ofcom launches an investigation into TikTok’s child safety measures, a growing chorus of experts is questioning whether outright bans on social media for teenagers are the answer. Psychologist Candice Odgers, who has studied adolescent mental health for 25 years, argues that the current debate obscures deeper issues—from the lingering effects of the pandemic to the mental health of adult caregivers. "The quickest way to make the internet safer for kids would be to kick all adult men off it," she told The Guardian, only half-joking. "But that’s not a policy we can adopt."

Odgers’ research suggests that the relationship between social media use and mental health is far more nuanced than often portrayed. While excessive screen time can exacerbate anxiety and depression, for many teenagers—particularly those from marginalised communities—online spaces provide vital support networks. "For some kids, social media is a lifeline," she said. "The idea that we can just ban it and make the problems go away is dangerously simplistic."

The UK’s Online Safety Act, which came into force last year, takes a more targeted approach, requiring platforms to protect children from harmful content while preserving access to beneficial online spaces. But enforcement remains a challenge. Ofcom’s investigation into TikTok, which could result in fines of up to £18 million or 10% of global revenue, underscores the difficulties of regulating a global industry with rapidly evolving risks. As one child safety advocate put it, "We’re playing whack-a-mole with harms that mutate faster than the law can keep up."

What to watch

The coming months will test the UK’s ability to balance innovation with responsibility. Key developments to monitor:

• Gene therapy rollout: The NHS’s ability to deliver SMA treatments at scale will set a precedent for future genetic medicine programmes.
• AI regulation: The government’s response to the Australian safety warning could signal whether the UK will adopt a more assertive stance on AI governance.
• Datacentre policy: With energy demands rising, expect pressure on the government to introduce sustainability standards for the sector.
• Child safety enforcement: Ofcom’s TikTok investigation may reveal whether the Online Safety Act has teeth—or whether platforms will continue to prioritise growth over compliance.