Question to the Department for Science, Innovation & Technology:

To ask His Majesty's Government what assessment they have made of concerns that AI technologies are being introduced prematurely to customers before their potential risks can be fully assessed.

The AI Regulation White Paper, published 29 March 2023 set out a framework for regulating AI that seeks to balance the need to address risk and support innovation.

As part of the regulatory framework, the UK is proposing a range of central functions through which the government will monitor known and emerging risks as AI technologies evolve. This will support us in assessing the effectiveness of our framework to address AI risks, and identify gaps in our risk mitigation efforts. For example, we are creating a horizon scanning function and a central risk function which will enable the government and regulators to monitor future risks, including ‘high impact but low probability’ risks such as existential risks, or AI biosecurity risks, in a rigorous, coherent and balanced way.

Tools for trustworthy AI - including internationally developed standards, and assurance techniques - will play a central role in the implementation of the framework, especially for those technologies already being introduced to the market. Through AI assurance, businesses and consumers are better able to decide whether a product or service using AI is legitimate and trustworthy. Impact assessments, performance testing and, possibly, pre-release verification or certification against AI standards in the longer term, are a few of the assurance mechanisms that can help organisations innovate responsibly while also determining whether an AI system complies with applicable standards and regulations.

The collaborative, adaptable framework outlined in the AI regulation white paper will use the proposed central functions to convene and learn from the expertise of frontier researchers, industry, academics, representatives of the public and other key stakeholders as we continue to develop policy in this evolving area.