'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

British lawmakers and OpenAI have reached a consensus that artificial intelligence (AI) is too significant and potentially hazardous to govern without a solid legal framework. The UK has historically advocated for a flexible, middle-ground strategy to regulate AI by utilizing existing regulators and voluntary agreements from developers.

However, the Joint Committee on Human Rights has now released a comprehensive report advocating for an extensive bill to cover AI, ensuring transparency and compliance through an independent regulator. Over 70 MPs and peers have separately expressed support for halting the development and operation of AI in Britain. The proposed legislation would also grant monitoring and control powers over such systems, though the notion of artificial superintelligence remains theoretical rather than a reality.

Alex Sobel MP, who chairs the Joint Committee on Human Rights, emphasized in a statement that the UK, along with other nations, lacks a legislative and regulatory approach to AI that is adequate for the current situation. Both the Joint Committee on Human Rights and OpenAI contend that the existing voluntary disclosure system is insufficient.

OpenAI has released its own manifesto, advocating for national AI safety regulations, including independent assessments and rules for slowing or halting AI development. The company also believes that rules should prioritize the most prominent AI companies, not simply create a single rule for every small firm experimenting with a language model (LLM).

While OpenAI's interest in regulation is understandable, it is essential to acknowledge that independent assessments and rules for large AI companies are necessary. The human rights committee's report acknowledges the challenges posed by the complex AI supply chains, as responsibility can be dispersed among developers, deployers, and users.

The government has acknowledged the need for review, but the report underscores the urgency of taking action. Britain does not require regulations for every chatbot, but it should not delay in establishing rules. A predictable AI regime would benefit the economy by providing companies with clear guidelines and discouraging reckless behavior.

Nonetheless, the voluntary governance approach presents a contradiction, as the organizations developing frontier AI are expected to determine the appropriate risk levels for society while facing immense financial investments. OpenAI's support is politically advantageous, but it should not grant the company excessive influence over regulatory decisions.

Ultimately, it is clear that the combination of lawmakers, researchers, and OpenAI's agreement that voluntary commitments are insufficient necessitates a more robust regulatory framework. The lawmakers must recognize that continuing to rely primarily on voluntary commitments is no longer a viable option.