4: Data and Artificial Intelligence

Estimated read time: 2 minutes

When OECD Digital Government Reviews reflect on data and AI they do so using in-country interviews, organisational surveys, peer insights, the OECD Framework for the Data-Driven Public Sector, the OECD Good Practice Principles for Data Ethics in the Public Sector, and the comparative data drawn from the relevant parts of the OECD Digital Government Index. That Framework analyses the three pillars of 1) data governance, 2) applying data to generate value and 3) data for trust which collectively allow for countries to develop their data maturity.

The conversation about data in the UK government has often been a maelstrom—full of promise but fraught with challenges. While there are pockets of excellence, there’s an overarching issue with gaps and inconsistencies that suggest a significant reset is needed. When it comes to AI there is a strong track record over time of doing important things to create the right foundations, as well as a no doubt somewhat distracting atmosphere of enthusiasm (if not hype) from external vendors, government advisers and political leaders. To truly move towards a data-driven, AI-enabled public sector that can effectively and responsibly maximise social good, the UK must address foundational elements of data governance, unlock the value of data, and build trust with the public.

As with other parts of this review I am not drawing on the extensive interviews, surveys and comparative data which the OECD team would access in conducting a similar exercise. As a result, some of these opinions might be better informed than others but all of them are my own, and my own only, so where I am drawing the wrong conclusion or making the wrong point I look forward to being corrected. Some of that might be subjectively wrong, some of it objectively so – in either event please educate me, and in particular share excellent examples where existing strategy, practice and impact are demonstrably having a fantastic impact that I am unaware of and have overlooked.