Agenda item

Algorithm Register

Minutes:

The Chairman invited Cllr Neil Prior and Juliet Whitworth, Head of Research and Information, to introduce the report.

 

Juliet updated members on the discovery work which officers had been undertaking to ascertain the level of activity currently underway in this field. The benefits of establishing an Algorithm Register were described including the sharing of best practice and providing open and transparent information for citizens. The furthest advanced project in the UK is being led by the Cabinet Office in partnership with the Ada Lovelace Institute, the Alan Turing Institute, the Committee on Standards in Public Life, the Open Government Partnership, the AI Now Institute and the think tank, Reform. Lead members had agreed that the LGA should join this project at an early stage to enable greater input and influence over its direction, including a leading role for local government. Over the coming months, officers would be promoting the use of predictive analytics with member councils with a view to piloting some projects with LGA support.

 

Cllr Prior reiterated his support for the decision to join the Cabinet Office register.

 

Members’ comments and questions:

·       It was suggested that officers look to Estonia, who are well advanced down this path already.

·       In response to a query about possible duplication with LG Inform, Juliet confirmed that data for predictive analytics would generally be more granular than that from LG Inform and so both could continue to be used side by side.

·       Members expressed some caution about bad publicity that algorithms had received in the press in recent times and so welcomed the Government taking the lead on a register. They also considered it would be helpful to be able to demonstrate real life examples of where this was making a difference to residents and communities.

 

Decision:

Improvement & Innovation Board noted the update and endorsed the decision by Lead Members to join the project to create a government algorithm register.

 

 

Supporting documents: