Use of AI in Government

The Public Accounts Committee have launched an inquiry regarding the use of AI in government.

The aim of this inquiry is for the public sector to set an example in the safe and ethical deployment of AI, which both has the potential to transform services for the taxpayer and presents risks and concerns.

Government has begun to experience an AI revolution, where AI capabilities increase the potential for efficiency gains. However, AI, in all its encompassing forms, is not infallible and must be applied diligently and appropriately. Much like the data/digital transformation that the government experienced in the 2010s, they must now use that learning to ensure a smooth and just transition to AI and its services.

The IET recommendations are:

• Strong data foundations will allow government departments and data professionals to utilise AI effectively and execute their job well.
• Competency is pivotal to the successful uptake of AI. To encourage private sector collaboration, the public sector needs to show competency and operability through regulation and skills.
• AI ethics must be central to driving the use of AI in government. Ethics frameworks should be updated to include the impact of a variety of AI products and to consider users of AI. 
• Strategic application of AI is key if government is going to successfully adopt AI. Although there are many facets of AI to invest in, government needs to invest effectively and strategically if there is to be successful implementation of AI. 
• Industry standards should not only aim to be met but exceeded. The EU AI act, along with the professional standards published by the ISO, provide a good basis for the deployment of AI, these should be consulted and applied where appropriate.
• Sustainability is a major concern in relation to AI. AI regulations should also look beyond the immediate risks of its development to the broader impact it has on the environment. 
• Searchability: Data and information must be able to be searched and found with options in the metadata to enhance usability and transparency.
• Cybersecurity must include the protection of the data, access to data and AI services, where the algorithm is going to be hosted, and ISO cyber standards.