Supporting the adoption of AI in government statistics
Artificial Intelligence (AI) has become increasingly important across all parts of the Analysis Function and the Civil Service. The common question that we hear isn’t whether we want AI to be embedded into our work, it’s to what extent AI can be safely used and what opportunities there are.
We’ve already seen examples of use, from coding to helping teams manage, process and understand information more efficiently. However, the technology is still maturing and the analytical community are still learning how to use AI for our day-to-day work, and how to ensure that it adds value while being used responsibly and ethically. For the Government Statistical Service (GSS), the principles of Trustworthiness, Quality and Value form the core of our approach, and that applies to AI use too.
Action in the GSS
The GSS AI Steering group was set up at the beginning of 2026, bringing senior leaders across the statistical system together with a focus on four key objectives:
- to understand the current AI landscape within the GSS
- to identify and address gaps in AI use and guidance for statistics
- to establish mechanisms for sharing and updating best practice
- to identify any actions needed to develop AI capability within the GSS
The group has been linking up with AI initiatives across government, including the Analysis Function AI group and the Royal Statistical Society (RSS).
Recently, a complementary GSS AI Task and Finish Group has been established to help the statistical community share learning, identify good practice, and support the responsible adoption of AI across government statistics. This task and finish group will focus on practical outputs that help statisticians across government to understand how AI can be applied effectively and safely. Their workstreams are:
Case studies
Lead by Adrien Speight, Department for Transport
We want to collect and share practical examples of AI use across the GSS. The group is looking for examples of positive stories, but also challenges and lessons learned. These will help highlight where some of the pitfalls can occur. A standard template is being developed and the group will be working with the GSS Heads of Profession to gather examples from across departments. These will be shared through GSS communications, including a seminar programme and wider community channels, helping colleagues learn from successes and challenges across government.
Seminar series
Lead by Laurynas Vasys, Department for Science, Innovation and Technology
A programme of events is being developed to showcase AI in practice, bringing real-world examples to life while helping statisticians to build their knowledge, skills and confidence in using AI. Themes will likely explore ethical, effective and transparent approaches to AI use for statisticians and official statistics.
Guidance and good practice
Lead by Sadi Md Akmal Hossain, Department for Work and Pensions
We are reviewing existing guidance and identifying opportunities to strengthen support for statisticians using AI, while aligning with wider government work on assurance, governance and responsible use.
What’s next? Getting involved
We would love to hear from colleagues at all stages of their AI journey, whether you are exploring ideas, running pilots or already using AI in production. If your team has a case study, lesson learned or good practice to share, please respond to our upcoming call for contributions through your Head of Profession and departmental networks.
The more experiences we can bring together from across the GSS, the stronger our collective understanding of how AI can support high-quality, trustworthy statistics.
Weichao Wang, Trade Remedies Authority, and Rachel Skentelbery, Office for National Statistics