UK consultation: A pro-innovation approach to AI regulation

Updated April 30, 2024

 

In February 2024, the UK government published A pro-innovation approach to AI regulation: government response, which resulted from a consultation published in March 2023 (and updated in August 2023).

 

The "pro-innovation approach" of this consultation and response is substantially less regulatory and more market-oriented than that of the EU AI Act and arguably less regulatory even than that of the US Executive Order on AI. There have been some recent reports (paywall) that the UK is reconsidering this very light-touch regulatory approach.

 

The consultation response applies a “context-based” approach to AI regulation (i.e. applying regulation in sectoral contexts on a technology-neutral basis) with five cross-sectoral principles:

  • Safety, security and robustness;

  • Appropriate transparency and explainability;

  • Fairness;

  • Accountability and governance; and

  • Contestability and redress.

 

For the time being, the UK government will apply this high-level approach on a non-statutory basis. However, the response does indicate willingness of the UK government to take regulatory action in appropriate circumstances, including:

  • noting with approval ongoing UK regulatory actions involving AI, including from the Competition and Markets Authority review of foundation models and the Information Commissioner’s Office guidance on data protection and AI (para. 14)
  • suggesting that where "current legal frameworks and regulator remits may not effectively mitigate the risks posed by highly capable general-purpose AI systems ... new responsibilities on the developers of highly capable general-purpose models may more effectively address risks" (para. 65)
  • setting out an AI regulation roadmap for 2024 (para. 93).

 

The consultation response summarises various aspects of the evolving UK regulatory approach, including:

  • cooperation initiatives internationally, within the UK government (including through the Digital Regulation Cooperation Forum) and with the private sector
  • funding for AI initiatives, including £10 million to build UK regulators' AI capabilities
  • developing a "cross-economy AI risk register", which is highly aligned with our "harms register" approach on Saihub.info
  • technical / governance steps for companies proposed by the UK Department for Science, Innovation and Technology in October 2023.