Scotland is to become the first part of the UK to make it mandatory to register the use of AI tools or applications within the public sector.
Every public sector project using AI will be recorded on the Scottish AI Register, a publicly-accessible database which aims to provide citizens with information about the use of the technology.
The register is currently voluntary, but mandatory registration will be gradually phased in, starting with Scottish government departments.
Richard Lochhead, innovation minister for the Scottish Government, said the scheme aims to improve trust and bolster confidence in the use of the technology across the public sector.
"From cancer diagnostics to helping our net zero journey, AI is a powerful and rapidly-developing tool the public sector can use to help drive efficiency and deliver solutions," he said.
"Making it mandatory for public sector use of AI to be registered will not only give the public increased confidence that AI is being used openly and transparently, but will also act as an increasingly powerful source of best practice, helping ensure AI is used in ways which is both economically and technically viable and makes a positive impact across society."
The register will include information on the development and use of each AI system deployed by a public service or government department, from conception through development into operation, and, when applicable, retirement.
This will include details on the data used in any AI system, how the technology works, and information about risk management procedures and accountability.
As part of the scheme, members of the public can also request information on the use of the technology and provide departments or services with feedback.
One organization that's already listed as a project is Scottish Children's Reporter Administration, which is exploring several potential theoretical ways that AI could be used within the Children’s Hearings system.
"As a public body working in a sensitive area of service delivery we wanted to be fully up front and open about our early exploratory work on potential future uses of technology that might have a positive impact on our skilled work," said chief executive and principal reporter Neil Hunter.
"Our involvement in the register also unlocked a lot of support and advice from across Scottish Government and partners on issues of research and evidence, experience elsewhere from a national and global perspective – and most critically for us – access to expertise on issues of ethics, impact, rights and privacy."
The move from the devolved Scottish administration comes as the UK government hopes to integrate AI tools widely across public services.
Last year, it announced the creation of a new ‘Incubator for AI’ team and vowed to upskill thousands of civil servants in areas such as programming, engineering, data science, and machine learning.
"The potential productivity benefits from applying these technologies to routine tasks across the public sector are estimated to be worth billions," said deputy prime minister Oliver Dowden at the time.
The use of AI and emerging technologies in the public sector has been a contentious topic over the last 18 months amid the generative AI boom.
Data privacy and public safety have been flagged as key concerns by industry stakeholders and think tanks, with research in February showing that public sector bodies aren’t always rigorous in their use of AI.
A Freedom of Information (FoI) request by digital adoption firm WalkMe showed that only one-quarter of UK public bodies have established safe use policies for AI in the workplace.
While 19% are in the process of developing policies, 39% currently allow employees to use the technology without having any safeguards in place.
