"AI adoption is very different to previous tech shifts in the media" – Q&A with expert Huw Davies

1 Business applications of Artificial Intelligence “ AI adoption is very different to previous tech shifts in the media” – Q&A with expert Huw Davies – Employer Apprenticeship Trailblazer Chair for Data Engineer, ML & AI Standards [Skills England] and BBC Technology Operations Duty Manager AI Skills for Business Competency Framework Case Study Series 2 Business applications of Artificial Intelligence “ AI adoption is very different to previous tech shifts in the media.” Huw Davies, Employer Apprenticeship Trailblazer Chair for Data Engineer, ML & AI Standards [Skills England] and BBC Technology Operations Duty Manager Since 2014 Huw Davies has played a central role in the UK working with employer consortiums, educators and sector skills bodies spearheading the development of apprenticeship Standards for broadcast, data, and machine learning (ML) engineering, as well as delivering one of the UKs most successful apprenticeship Standard for artificial intelligence (AI). These levy funded apprenticeship Standards have enabled UK employers and Apprenticeship Providers to deliver critical in-demand training to upskill “UK plc”; following a colourful career working across technology and a decade leading a portfolio of early career and apprenticeship training programmes at the BBC. In this Q&A, he reflects on the importance of aligning technical standards with real-world needs, how frameworks such as the AI Skills for Business Competency Framework can support responsible adoption, how organisations like The Alan Turing Institute can act as trusted, impartial sources of information, and what it takes to build AI capability across a complex public service organisation. Key takeaways • AI adoption is happening quickly within organisations – the required skills are no longer niche, and every department needs tailored capabilities to adopt AI successfully and responsibly. • Apprenticeships have played a key role in meeting demand for data and AI skills, and they need to keep evolving to match the pace of change. • The AI skills ecosystem can feel fragmented – better coordination across frameworks and standards bodies is essential to support learners and employers. • The AI Skills for Business Competency Framework offers a clear, accessible, trusted reference point for employers building AI capability. • Embedding AI capability across large, complex organisations may require a decentralised model, as well as shared understanding and consistent skills. • As AI adoption increases, professional chartership and regulation may become key ways to ensure competence and trust. 2 Q&A with Huw Davies, BBC, BBC Technology Operations Duty Manager 3 Business applications of Artificial Intelligence Can you start by introducing yourself and your work in relation to AI skills? Huw: I’m Huw Davies. I’ve been involved in apprenticeship strategy and skills development for quite a while now – particularly with Skills England and in my previous L&D Schemes Manager role at the BBC, where today I work these days as a Duty Operations Manager for Technology Operations, helping to protect and keep the organisation’s critical technical infrastructure and BBC services up and running for our people & global audiences. I have led the development of several key technical standards, including the Level 7 Artificial Intelligence (AI) Data Specialist apprenticeship, which we first started working on in 2017 and published in 2019. That followed earlier work on broadcast engineering and technical operations-apprenticeships, which was a response to a critical skills gap in the context of an ageing workforce and the shift to cloud-based production and media tools for UK broadcasters and media organisations. AI is now gradually being mapped on to those broadcast engineering standards, too. Since 2021 we went on to develop Level 5 Data Engineer and Level 6 Machine Learning Engineer apprenticeships which were published last year. That whole progression pathway is now available for employers and providers, and it came out of real demand from across sectors – organisations were saying there was a gap when it came to AI, data engineering, and machine learning roles. 3 Q&A with Huw Davies, BBC, BBC Technology Operations Duty Manager 4 Business applications of Artificial Intelligence Why do you think AI skills are so important right now – particularly in the media and the wider workforce? Huw: It feels like we’re in the middle of a significant shift. I’m part of the first generation in my organisation that’s interacting with a machine and being given an AI assistant tool to work with. And there’s now an expectation that I’ll look at how I work and how AI might help me automate parts of that. That’s new – and it’s happening fast. It feels very different to previous tech shifts in the media. The move from analogue to digital media was gradual. Cloud and streaming changed the game, but this is another level – especially in the last 18 months. Tools like ChatGPT, Gemini and Copilot are already in use across businesses, sometimes informally and sometimes without oversight. The BBC has taken a very thoughtful approach with a clear policy, which I think is essential. But more broadly, every part of an organisation will need different AI capabilities – whether it’s HR and recruitment, finance, content production, or journalism. The skills aren’t just technical either. It’s about being able to implement safely, responsibly, and with trust. “ I’m part of the first generation in my organisation that’s interacting with a machine and being given an AI assistant tool to work with. That’s new – and it’s happening fast.” 4 Q&A with Huw Davies, BBC, BBC Technology Operations Duty Manager 5 Business applications of Artificial Intelligence How have skills frameworks helped shape your approach to AI training and apprenticeships? Huw: I’ve always used frameworks as a way to ensure quality and consistency – even back in my engineering management days, we relied on the UK-SPEC framework from the Engineering Council to benchmark skills together with support from organisations like the Institute of Engineering & Technology (IET). During the development of early AI and data standards, there weren’t many frameworks we could lean on, especially ones focused on AI. Since then, the AI Skills for Business Competency Framework has been a really helpful tool. We use it as a reference point – particularly with employers – to think about what kinds of skills are needed, and how to develop capability across different roles. It’s not overly technical, which makes it accessible to leaders and strategic decision-makers. That said, we often pair it with more detailed engineering frameworks when we need that finer granularity. How has the AI Skills for Business Competency Framework supported your work on apprenticeships? Huw: One of the challenges we’ve faced in developing AI-related standards is the lack of shared definitions across roles – especially when it comes to distinguishing between different data science and AI specialisms. That makes it harder to set clear expectations and build consistent training pathways. The Alan Turing Institute has played a really valuable role as a trusted, impartial convenor in this space. They gave us feedback on the Level 6 Machine Learning Engineer apprenticeship using the AI Skills for Business Competency Framework, which helped clarify expectations and led to changes in the standards during the approvals process. It’s a good example of how frameworks like this can support both quality and consistency in the system. What are your reflections on working with multiple frameworks in the AI skills space? Huw: From a user’s point of view, it can feel fragmented. Many organisations offer standards, guidance and frameworks, including Professional, Statutory and Regulatory Bodies (PSRBs), trade bodies and private industry. Each has their own approach and products, and sometimes they may be competing with each other. That makes it hard to see the links between them. We’ve seen how difficult it can be when different bodies don’t align or work together. What we really need is collaboration – a kind of metaframework that brings it all together. It’s not about replacing existing standards, but making sure they interoperate. That way, employers, training providers, and individuals aren’t left trying to piece it all together themselves. The AI Skills for Business Competency Framework is in a great position to be that glue – particularly because it’s not tied to a commercial product. It’s a trusted, independent reference point, and that’s what makes it so valuable. 5 Q&A with Huw Davies, BBC, BBC Technology Operations Duty Manager 6 Business applications of Artificial Intelligence Can you share an example of how AI capability has been enhanced at the BBC? Huw: One great example is the data-driven journalism apprenticeship I created at the BBC in response to the challenges of the pandemic. During COVID, I observed journalists were suddenly being handed raw data – often in Excel spreadsheets – and were expected to make sense of it quickly. Many of them had little experience with spreadsheets, would seek support from IT support helpdesks and our limited pool of Data Journalists we had at the time; journalists generally work using text-based systems, word processing their stories and are often in some instances on the road working from their mobile phones. We had to react fast to empower them to have the confidence, literacy and capability to understand data and produce accurate and trusted news content for our local and global audiences. So, working with a university and a private training provider, I adapted the Level 3 Data Technician apprenticeship – which many providers were already using in an AI context – and turned it into something specific for journalism within a period of around six months. It’s had a real impact; the BBC Academy has now put more than 200 journalists through that training across the UK, and it was designed to be open and accessible to journalists and news researchers from other news organisations benefiting the wider media industry. What should organisations be thinking about when building AI capability more broadly? Huw: You need to look beyond literacy. Yes, people need to understand AI at a basic level, but they also need access to the right people – whether that’s analysts, data engineers, machine learning engineers, or AI architects and product managers – depending on what they’re trying to do. For some organisations, that will mean learning how to source their AI capability from outside, while others will need to learn how to develop the expertise in-house. In organisations like the BBC that are large and complex, we’re increasingly seeing a hybrid operating model emerge as we experiment, develop and release trusted AI solutions across the business or to our audiences. Rather than only relying solely on our centralised AI group made up by senior leaders from editorial policy, strategy, data protection, information security, commercial, legal and technology teams who provide robust central governance and policy direction. This feeds into our divisional and embedded AI teams across individual departments. These emerging decentralised operating models enable tailored innovation while maintaining strategic alignment to our AI Principles. However, they also highlight the critical need for consistent AI literacy for our people, shared standards, and a unified understanding of ethical and practical applications across the organisation. The AI Competency Framework helps businesses bridge into decentralised AI operating model, guiding organisations providing a structured, role-based approach to have consistent AI literacy across the business, and enabling the tailored upskilling of new and existing staff enhancing the organisations AI capability. Transparency, explainability, and trust are all crucial. Once you lose trust, especially as a public service broadcaster, it’s incredibly hard to get it back. 6 Q&A with Huw Davies, BBC, BBC Technology Operations Duty Manager 7 Business applications of Artificial Intelligence Looking ahead, what are the priorities for the AI skills system in the UK? Huw: I think the apprenticeship standards we’ve developed are strong and have landed at the right time. But we’ve got to keep moving. AI is evolving quickly, and the skills system needs to adapt with it. We need to make sure apprenticeship training is efficient and responsive – but also open and accessible. Our Trailblazer Group have already spent the last year updating the L7 AI Standard we published in 2019. So much has changed, and I hope to seek Skills England approval to re-publish this towards the end of this year; including making it more accessible to a younger audience to fast-track eligible apprentices from the age of 18+ into highly specialist AI occupations. There’s definitely room for a next step – maybe a more junior AI apprenticeship, or something that focuses on roles like AI architects and product managers. And we also need to ensure AI is being sprinkled throughout all apprenticeship standards, not just the technical ones. Every profession is going to be touched by this. We’re also moving towards greater regulation in AI, and professional chartership could become one way of demonstrating competence and accountability. We have it in HR, finance, and so on, so why shouldn’t we have it in AI? Above all, we need to keep the focus on capability and not just qualifications. Frameworks like the AI Skills for Business Competency Framework are a great way to start that conversation, but we also need to join the dots across the whole system. “ The AI Skills for Business Competency Framework isn’t tied to a commercial product – that’s what makes it so valuable. It’s a trusted, independent reference point.” 7 Q&A with Huw Davies, BBC, BBC Technology Operations Duty Manager 8 Business applications of Artificial Intelligence This case study is published under the AI Skills for Business Competency Framework workstream. The framework sets out the high-level competencies required to ensure the safe and responsible introduction of Artificial Intelligence (AI) to organisations from the general workforce, through to AI specialists and leadership. It is sponsored by the Department for Science, Innovation and Technology (DSIT) and is funded by the Innovate UK BridgeAI programme, which empowers UK businesses in high-growth sectors, driving productivity and economic growth through the adoption of AI. This work is supported by a consortium including Innovate UK, Digital Catapult, The Alan Turing Institute, STFC Hartree Centre and BSI. Since its launch in May 2024, the Framework been adopted across public, private and government organisations in the UK. We would like to extend our gratitude to Huw Davies, Employer Apprenticeship Trailblazer Chair for Data Engineer, ML & AI Standards [Skills England] and BBC Technology Operations Duty Manager, for his significant contributions to the development of this case study. Special thanks to Alexandra Araujo Alvarez, Senior Research Community Manager for BridgeAI, Dominica D’Arcangelo, Programme Manager, and Marie Chappell, Programme Coordinator, from the Alan Turing Institute for their leadership and support. We also thank Stuart Gillespie for his role as the technical writer for this and other case studies in the programme. This work is led by Dr. Matthew Forshaw (Senior Advisor for Skills) and Dr. Clementina Ramirez (Skills Officer). Dr. Vera Matser (Head of Strategic Capabilities) is Principal Investigator for BridgeAI at The Alan Turing Institute. For any comments, questions, or collaboration opportunities with BridgeAI, please email: [email protected]. Acknowledgement Authors and Contributors Stuart Gillespie: Technical Writer (0009-0003-1402-0184) Huw Davies: BBC, BBC Technology Operations Duty Manager Matthew Forshaw: The Alan Turing Institute, Senior Advisor for Skills (0000-0001-7014-9837) Clementina Ramirez: The Alan Turing Institute, Skills Officer (0000-0002-6643-8828) This publication is shared under CC-BY 4.0 License on Zenodo. DOI: 10.5281/zenodo.17435041 8 Q&A with Huw Davies, BBC, BBC Technology Operations Duty Manager 9 Business applications of Artificial Intelligence Learn more at bridgeai.net