Government launches guidelines for AI procurement

The UK government has launched a set of guidelines to inform public sector procurement for artificial intelligence (AI) technologies.

Created in collaboration between the government’s Office for AI and the World Economic Forum Centre for the Fourth Industrial Revolution, the guidelines follow the publication of the Data Ethics Framework and the Guide to AI and are intended to enable public bodies to buy and adopt AI tools “confidently and responsibly”.

A pilot of the guidelines will start this autumn and the “living document” will be updated as the technology evolves to incorporate new lessons and practices from the various stakeholders involved in procurement processes.

Issues that the guidelines aim to tackle include the fast-moving and immature nature of AI technology and the lack of standards around technical specifications and considerations around ethics, which make AI systems more difficult to procure.

The document is structured around the main commercial stages of procurement, with emphasis on the preparation and planning stages.

Exploring the challenge rather than specific AI systems is one of the main areas covered by the guidelines, as is the definition of the public benefit involved in using AI while assessing risk.

The inclusion of procurement processes within a strategy for AI adoption and the consideration of technical feasibility around obtaining data are also part of the document, which advises buyers to consider aligning their work with relevant AI-related initiatives from government.

Having diverse, multidisciplinary teams is also part of the report, as is the creation of a “level and fair playing field for suppliers” and consideration to the lifecycle management of AI systems.