Research and analysis

Annex A: Methodology

Published 29 October 2025

Applies to England

A qualitative, exploratory design was adopted to facilitate deep engagement with main stakeholder communities and surface evidence that reflects real-world practices, barriers, and innovation models. While the primary data collection for this report was qualitative and based on six national stakeholder workshops and a senior roundtable, the sectoral analysis additionally draws on desk research to contextualise findings. This included a review of recent industry reports, government publications, and peer-reviewed academic studies related to AI adoption and workforce skills development across priority sectors. These sources were used to supplement gaps in workshop coverage, provide statistical context, and align the report’s insights with broader national evidence on AI integration and training needs.

The workshops questions were reviewed by experts from Skills England and the British Academy, and changes were made based on their feedback. Workshop participants were selected using purposive sampling, based on:

• their expertise
• their professional roles
• their relevance to one or more of the 6 stakeholder groups identified as central to the research

All invited experts were actively involved in AI skills development, workforce inclusion, or strategic planning. Care was taken to include voices from both highly digitised environments and those where AI adoption remains nascent or uneven.

A total of 43 organisations were represented across the six workshops and experts from further organisations participated in the roundtable.

Education and training organisations included Sixth Form Colleges Association, Ada National College, HOLEX, universities such as Royal Holloway, University of London, University of Surrey, University of Liverpool, Anglia Ruskin University, multiple FE colleges, and adult learning hubs, Association of Employment and Learning Providers (AELP), Baker Dearing Educational Trust, and SKEMA Business School.

Public sector and Social Care providers included NHS, Regional Directors of Adult Social Services, ADASS West Midlands, and local authority staff.

Industry and employer bodies included LinkedIn, Google, Institute of Directors, Bloomberg, Skills Provider Networks, SMEs, TechUK, Vertis Media, and JMAC.

Charities and inclusion-oriented organisations included Good Things Foundation, Shaw Trust, Latin American Disabled People’s Project, and Asian Resource Centre of Croydon.

AI training providers and consultancies included Educate Ventures, AI Confident, Institute for the Future of Work, The Alan Turing Institute, Surrey AI Centre and other technology-led upskilling initiatives providers, Tesseract Academy, Taught by Humans, Surrey Institute of People-Centred AI, and AI in Justice Programme.

Policy and research organisations included the Department for Education, The British Academy, Ofcom, King’s Trust, and Edge Foundation.

The inclusion of both national and hyper-local stakeholders allowed for the identification of macro-level challenges and community-level solutions, enabling insights relevant to both system-wide and regional policy considerations. Ethical approval was granted by Royal Holloway, University of London. Experts were given detailed information about the research, and informed consent was collected prior to participation. All data was anonymised, stored securely, and processed in accordance with GDPR and ethical standards for qualitative research.

Data was collected through a series of 6 thematic workshops and the senior stakeholder roundtable held between March 2025 to May 2025. The workshops were designed to explore stakeholder perspectives on AI adoption, emerging skills needs, access to training, and inclusion challenges across a wide range of sectors and communities. Workshops were structured to allow for open dialogue, critical reflection, and the co-creation of ideas, with each exploring various sectoral or regional themes. All workshops were recorded (with consent), transcribed, and analysed using an iterative, reflexive thematic analysis approach using Braun and Clarke (2012) and Braun and Clarke (2021) approaches. This method was selected for its ability to support rich, grounded interpretation of complex qualitative data across diverse stakeholder perspectives. The analysis followed Braun and Clarke’s widely recognised six-phase model.

Following the workshops, a senior stakeholder roundtable was convened at The British Academy in May 2025 to refine the outputs, including the draft AI Skills Framework and Employer AI Adoption Checklist. This roundtable session enabled testing of findings, validation of relevance, and clarification of sector-specific considerations.

Participants included representatives from:

• Skills England
• Department for Education (DfE)
• Department for Science, Innovation and Technology (DSIT)
• Department for Work and Pensions (DWP)
• Ofcom
• LinkedIn
• Google
• Cabinet Office
• British Academy
• BCS, The Chartered Institute for IT
• Anglia Ruskin University
• The Good Things Foundation
• Sixth Form Colleges Association
• Skills Builder
• The King’s Trust
• Royal Holloway, University of London

Their feedback informed further refinements and helped ensure that outputs reflected both strategic and operational priorities in AI workforce planning.

Following the workshops and senior stakeholder roundtable, further feedback was gathered through a series of presentations and experts’ reviews where the proposed AI Skills Framework, Employer Adoption Pathway Model, and Employer AI Adoption Checklist were tested with diverse audiences. This feedback was used to refine the outputs and ensure their relevance across different sectors and stakeholder groups.