‘Nowhere in the world, including the UK, has a current legislative and regulatory approach to AI that is fit for purpose’ — Britain’s lawmakers have started calling for AI regulation, and they should refuse to settle for half-measures

• September 15, 2026

British lawmakers and OpenAI have found something they can agree on: artificial intelligence has become too important and potentially harmful to govern without a real legal framework.

The country has long promoted what it deemed a flexible, middle-ground approach to regulating AI through a mix of existing regulators and voluntary agreements by developers. But now, Parliament’s Joint Committee on Human Rights has issued a long and detailed report calling for an expansive bill covering AI, enforcing transparency and compliance via an independent regulator.

The pressure goes considerably further than ordinary complaints about AI chatbots. More than 70 MPs and peers have separately backed calls for Britain to prohibit the development and operation of AI. The legislation would also create monitoring and control powers over such systems, although artificial superintelligence remains hypothetical rather than something currently sitting in a server farm plotting its next move.

“AI is heralded as an unprecedented era of technological development with the potential to transform our lives for better or for worse. It is moving with such speed and complexity that its impact is hard to accurately predict. What is clear is that at present we are unprepared to deal with its consequences however potentially dire they may be,” Chair of the Joint Committee on Human Rights Alex Sobel MP said in a statement.

“Nowhere in the world, including the UK, has a current legislative and regulatory approach to AI that is fit for purpose. New legislation is needed to establish a comprehensive set of protections that deal with the entire AI supply chain and its lifecycle.”

OpenAI has become an unexpected ally for this kind of regulation. As one of the companies with the most to lose from badly designed AI regulation, OpenAI now says governments should start imposing mandatory safety requirements on frontier AI companies.

Protecting people means AI laws need consequences

The mostly voluntary, good-faith disclosure system is insufficient, according to both MPs and OpenAI. The company issued a manifesto of its own arguing for national AI safety regulation, including independent assessments and rules for when AI development should slow or stop. It also argued that the rules should focus on the biggest AI companies, not just make one rule for every small firm experimenting with an LLM.

There is an obvious self-interested element here. Regulation aimed specifically at the richest frontier labs would affect OpenAI, but sophisticated compliance regimes can also strengthen the position of companies wealthy enough to comply with them. A requirement for expensive independent testing is considerably easier to absorb when billions of dollars are sloshing around the balance sheet.

That does not make OpenAI wrong that regulation should follow capability rather than apply equally to every AI company. The company has also argued that democratically accountable standards and independent verification would be preferable to the current situation in which frontier laboratories largely decide their own safety rules.

That last point should be printed in very large type and pinned somewhere in Whitehall. AI companies can employ excellent safety researchers and genuinely care about responsible development while still being terrible substitutes for governments. We do not usually allow pharmaceutical companies to decide privately whether their own medicines have been tested enough, then thank them for their voluntary commitment not to poison anybody.

UK AI safety

The UK isn’t starting totally from scratch. The country’s AI Security Institute was created to study and test advanced models and has worked with frontier developers including OpenAI. Yet Britain’s broader system continues to rely heavily on existing regulators and voluntary cooperation rather than a dedicated statutory regime for frontier AI.

But the agencies set to supervise a specific industry are poorly positioned to deal with expansive general-purpose AI models whose capabilities stretch across dozens. The human rights committee’s report recognizes this problem. It argues that AI supply chains complicate accountability because responsibility can be scattered among developers, deployers and users. The government has said it is reviewing the situation, but the report makes it clear that action is needed soon.

While Britain does not need to regulate every chatbot like its Skynet, it shouldn’t have to wait for absolute proof of catastrophe before establishing rules. It’s a benefit economically, too. Companies prefer knowing what the rules are to discovering them after an accident or legal case. A predictable AI regime would make Britain more attractive to serious AI developers while discouraging reckless behavior.

The biggest reason to act, though, is that voluntary governance contains an unavoidable contradiction. The laboratories developing frontier AI are being asked to decide how much risk society should tolerate from products they are spending enormous sums to build. Even with honorable intentions, that is too much authority to place inside a handful of companies.

OpenAI’s support, while politically useful, shouldn’t give it any extra influence, however. Parliament should be particularly wary of allowing the largest AI companies to design rules that conveniently turn their enormous resources into a regulatory moat against smaller competitors.

Still, when lawmakers, researchers and one of the world’s leading AI developers all agree that voluntary commitments are no longer enough, continuing to rely primarily on them begins to look like lawmakers are just dragging their feet.

“Fundamentally, this is about making sure that you, as an individual, know when AI is being used in the decisions that affect you,” Sobel said. “We also want to make sure that if something does go wrong then avenues of redress will be available. We need these protections in now, it cannot wait until fear human rights risks become reality.”

Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.

Source

British lawmakers and OpenAI have found something they can agree on: artificial intelligence has become too important and potentially harmful to govern without a real legal framework. The country has long promoted what it deemed a flexible, middle-ground approach to regulating AI through a mix of existing regulators and voluntary…