Question to the Department for Science, Innovation & Technology:

To ask the Secretary of State for Science, Innovation and Technology, whether her Department plans to take steps to help ensure that AI (a) algorithms and (b) systems are transparent, explainable, and free from bias or discrimination.

Our recently published white paper sets out a framework for regulating AI, which is underpinned by five cross-sectoral principles which will inform the responsible development and use of AI. These principles include ‘Appropriate transparency and explainability’ and ‘Fairness’.

As set out in the white paper, our iterative framework aims to ensure that AI systems are appropriately transparent and explainable, allowing for individuals, regulators and organisations to have access to appropriate information about an AI system, as well as to be able to interpret and understand the decision-making processes behind them. Together, appropriate transparency and explainability will help to drive trust and understanding of AI systems.

The white paper is subject to public consultation open until 21 June 2023.

In terms of AI systems used in delivering public services, the Government was one of the first in the world to implement an algorithmic transparency standard for use in public service delivery, allowing public sector organisations to provide clear information about algorithmic tools they use to support decisions, including why they are using them. The Government’s AI procurement guidelines also recommend that systems being procured should undergo an equality impact assessment in order to ensure that AI meets the needs of the diverse society it serves.