Research and analysis

Employer guide: What works for AI upskilling in the UK

Updated 27 July 2026

Applies to England

Acknowledgements

This report received support from The British Academy Policy-led Innovation Fellowship award IF24RBDS\240034 (referred to as ‘the fellowship’ within the report), in partnership with Skills England. Input from different areas and sectors helped shape the workshop designs and findings. The opinions expressed in this report are those of the author and do not necessarily reflect the views of the department.

1. Introduction

This guide forms part of the SKAI programme, Skills for AI: What Works for AI Upskilling in the UK. The programme translates national evidence on AI skills into 4 connected outputs: employer guide, supporting case studies, insight briefing based on research, and a research evidence, analysis and methodology report.

This guide sets out what effective AI upskilling looks like in practice for employers across the UK. It draws on evidence from workshops with employers and AI experts, a UK-wide employer survey, and detailed case studies to identify the training approaches that support confident, safe, and productive use of AI at work.

The guide focuses on practical training principles rather than specific tools or technologies. It reflects the reality that most organisations are at an early or intermediate stage of AI adoption, and that workforce capability is often developed informally. By bringing together evidence from different sectors and organisational contexts, it highlights the features of AI training that are most likely to be usable, inclusive, and sustainable.

This document is intended to support employers, including SMEs and organisations with limited internal capacity, to think clearly about how AI training can be designed and embedded into everyday work. 

It forms part of a wider set of outputs. The evidence behind this framework includes 23 workshops, 10 case studies and a survey of 536 responses. The underlying evidence, sector analysis, and methodological detail are presented in the research evidence, analysis and methodology report, while fuller examples of delivery models are provided in the supporting case studies report.

2. Why AI upskilling matters now

AI is already changing how organisations work. Generative AI tools are now widely used across sectors, and agentic AI systems that can act with a degree of autonomy are emerging. AI is already embedded in everyday work. Our survey findings show that over 44% of organisations report using AI tools daily, often through existing workplace systems. These developments are rapidly affecting job roles, demand for skills, and day-to-day tasks.

National policy ambitions highlight the importance of getting this right. The AI Opportunities Action Plan and the UK Industrial Strategy both aim to ensure that AI-driven growth leads to long-term economic benefits that are shared across the workforce. This requires not only investment in technology, but also investment in people and skills.

Workshop evidence shows consistent challenges in developing an AI strategy and workforce skills, providing effective training, and using AI safely and responsibly. Without targeted support, there is a risk that AI adoption will remain uneven and that existing workforce inequalities may widen.

For employers, this creates an immediate challenge: how to build the skills needed to use AI effectively, safely, and across the organisation.

Most organisations are yet to engage deeply with AI or are still at an early stage of AI adoption. Many are experimenting with AI tools, but few have embedded them into everyday work or developed structured approaches to workforce capability. Organisations need to move from initial engagement to sustained, scalable and effective use in practice.

Good, targeted training can help take that step. 

In this work, AI skills are defined as “the competencies and abilities required to develop, implement, manage, and interact with AI systems effectively”. They are grouped into 3 categories: technical, responsible/ethical and non-technical AI skills.

Most roles require a combination of these skills, rather than advanced technical expertise alone.

This guide builds on the AI Skills for the UK Workforce report by focusing on the next steps: how to design and deliver effective AI training in practice, based on evidence of what works across organisations.

3. What good AI training looks like (PRIMES)

When asked to rank what makes effective AI training in practice, survey respondents highlighted the following as being of primary importance:

1: Flexibility and accessibility

2: Practical and contextualised learning

3: Recognition and progression

4: Clear and aligned AI skills frameworks

At a secondary level of importance, respondents listed:

5: Ethics, governance, and assurance

6: Impact measurement and sustainability

7: Inclusivity, community, peer support, and cascade models

8: Leadership and organisational buy-in

9: Quality assurance and trusted provision

Findings from survey respondents, along with insights from the workshops and case studies, led to the development of the PRIMES framework presented here. While nearly all organisations report providing AI training (97%), many still identify key gaps in provision, particularly in flexibility (51%) and practical, contextualised learning (34%). This highlights a shift from access to effectiveness. This is reinforced by wider gaps identified by employers, including lack of clear AI skills frameworks (35%), limited focus on ethics and governance (29%), and weak leadership or organisational buy-in (22%).

Together, these findings highlight the need for a more structured and consistent approach to AI training design. The PRIMES framework sets out 6 principles of effective AI training:

• practical
• reachable
• integrated
• modular
• expandable
• sustainable

The PRIMES principles give a clear and practical guide to what ‘good quality’ AI training should look like, flexible enough for different sectors to adapt to their needs. They focus on how training is planned, delivered, and kept up to date, rather than promoting any specific tools or providers. 

This creates a shared reference point that can help Skills England, sector bodies, training providers, and other relevant organisations check quality and build trust in training programmes. It can support approved provider schemes, endorsement processes, or quality standards as AI tools and skills continue to change.

Criteria for meeting PRIMES principles of effective AI training

Practical criteria:

• training links clearly to real tasks and decisions in the learner’s role
• learners practise using AI tools in work-relevant scenarios
• training makes clear when AI should and should not be used
• training recognises and builds on existing informal or self-taught AI use within roles

Reachable criteria:

• training is accessible in time (duration and flexible timing), format and language
• learners are given paid time or protected time to take part
• materials are suitable for people with different levels of digital experience
• personalised support is available for learners who need it
• training design explicitly considers intersecting barriers (for example income, age, disability, gender) rather than treating learners as single categories
• training is designed to build confidence alongside skills, particularly for learners with prior exclusion from digital or technical education

Integrated criteria:

• training fits within existing systems so that workers find practical value and can immediately apply learning to their roles
• content aligns with relevant professional standards, qualification or skills frameworks
• training reflects sector-specific values and responsibilities, such as safety, accountability, or person-centred practice
• provision supports benchmarking, quality assurance, and clear recognition of learning outcomes
• employers and learners can identify training that is reliable, relevant to their work, and appropriate for their level of AI use
• training aligns with organisational data infrastructure and digital systems to ensure practical and safe application
• baseline AI training is mandatory before staff use workplace AI tools, particularly where AI use involves organisational data, confidential information, regulated activity, safety, or professional judgement
• training helps learners understand relevant next steps, including further learning, safer AI use, or AI-related responsibilities where appropriate

Modular criteria:

• training is broken into short, manageable units
• learners can complete training in different formats and at different speeds
• content allows learners to enter at different levels of training according to their AI skills and return when needed

Expandable criteria:

• training develops transferable AI skills that can be applied across different tools, systems, roles and organisations where possible
• core content remains consistent while allowing local or role-specific adaptation
• provision can support rapid growth in demand as AI use spreads

Sustainable criteria:

• training can be updated as AI tools and uses change
• there are plans to review and refresh learning after delivery
• providers monitor outcomes, including whether training improves confidence, quality of use, and decision-making
• responsible use of AI is embedded as a core and enduring capability that remains essential as AI tools evolve - this includes confidentiality, data protection, transparency, and human oversight
• learners are given opportunities to reflect on how AI is used in practice and identify further learning needs and progress towards more effective AI use

3.1 AI training must be practical

Successful AI training is practical and helps individuals use AI in their day-to-day work. It links learning to real tasks and real decisions. 

Training should teach individuals how to work effectively with AI systems, not just how to access tools. This includes understanding how AI fits into their role and when it should or should not be used. Training should also recognise and build on existing informal or self-taught AI use, rather than assuming a zero-starting point.

Practical AI upskilling training programmes prioritise hands-on learning that links directly to real tasks. This includes:

• scenario-based activities such as drafting AI-supported care plans
• small applied projects with feedback
• use-case libraries that show safe and practical uses
• time for reflection and repeat practice

Practical training should cover the 3 AI skills categories set out in the AI Skills for the UK Workforce report:

• technical AI skills (for example write structured prompts, use low-code AI platforms for automation)
• non-technical AI skills (for example use basic AI tools to complete routine tasks, test AI tools for application in your tasks)
• responsible and ethical AI skills (for example assess accuracy and appropriateness of AI for tasks, identify bias in AI outputs)

Strong programmes do not treat these skills as separate topics. Instead, they bring them together within practical learning activities. Learners use AI tools, interpret outputs, and apply judgement as part of the same task.

Case studies highlighting good practice on this principle include NHS primary care settings using AI in real tasks such as documentation, and 100 School delivering challenge-based training where participants created tangible work outputs, enabling learning to be applied directly in practice.

3.2 AI training must be reachable

AI training programmes should be reachable for all workforce groups. Survey findings highlight that barriers are not only structural but also practical and psychological. Employers report cost (42%) and limited availability (37%) as key challenges to participation.  This means actively removing practical barriers by providing:

• protected and paid time to take part
• low-cost, subsidised or internally provided training options where possible
• well-designed course materials including plain language materials and closed captions (subtitles to support learners with hearing impairments or those accessing content without audio)
• technologically accessible training including mobile-friendly delivery and low-tech alternatives

Reachable AI training recognises that confidence can also be a barrier. Coaching and mentoring from local champions can help learners who lack digital experience. Simple self-reflection tools can guide workers to identify how they would use AI and help build confidence. Other approaches include entry level modules and safe spaces to ask questions and test ideas. 

Designing training to be inclusive from the outset helps ensure everyone can benefit. If accessibility is not built in, training is likely to favour more confident digital users, widening skills gaps. Good training design recognises that learners may face overlapping barriers, such as age, disability, income, or gender, rather than treating them as a single group.

An example of ‘reachable’ training is demonstrated in the case study of the Good Things Foundation. It offers accessible, bite-sized AI learning designed for digitally excluded groups, supporting confidence alongside skills development.

3.3 AI training must be integrated

Effective AI training should be built into existing systems, roles, and ways of working rather than treated as a standalone activity. Training should align with the tools, processes and standards people already use, and integrate with organisational data and digital infrastructure to support safe and effective practice. It should also support clear progression pathways into AI-related roles and responsibilities.

Case studies in this report show this approach in practice. Organisations such as KPMG and Airbus embed AI training within existing systems, workflows, and governance structures.

Across sectors, stakeholders emphasised that training is most effective when it connects directly to the digital tools and processes staff work with every day. When training reflects real systems, such as care records, building models, data platforms, or creative tools, people can use what they learn straight away.

Training must integrate sector values, responsibilities and regulation. In areas like health, social care, and construction, this includes a strong focus on safety, accountability, and person-centred practice. When AI training supports these values, people are more likely to trust the tools and use them responsibly.

Integrating AI content with established training frameworks, such as apprenticeships, professional standards, and mandatory training, helps normalise AI use. This shifts AI adoption from informal experimentation to consistent, governed practice. 

Integration also enables benchmarking, shared standards and formal recognition including micro-credentials linked to qualifications, CPD credits or digital badges. This supports progression within roles and professions and helps employers compare training provision. 

Clear approval and quality checks are also important. Approved provider lists, sector endorsements, and trusted intermediaries help employers and learners find reliable training that meets agreed standards.

3.4 AI training must be modular 

Effective AI training should be flexible in both format and pace, fitting around different roles, workloads and workplaces rather than fixed schedules. This principle is reflected in one of the case studies in this report. Platforms such as LinkedIn Learning demonstrate this approach through structured, modular pathways that enable learners to progress at different speeds and levels.

Good AI training offers a choice of delivery methods such as:

• in-person sessions
• online learning
• hybrid models
• workplace-based training
• self-paced study

This flexibility allows employers to choose options that suit their staff and business needs.

Modular design is essential. Short, focused units make it easier for people to access, especially for SMEs and frontline services. In many settings, very short (30 to 90 minutes), stackable modules are more practical than longer courses.

Microlearning helps people build skills gradually and revisit content when needed. Modular training also allows learners to start at different points and return to content as their skills needs change. This approach works well for SMEs, freelancers, and frontline staff who have limited time.

Flexible training also supports different learning speeds. Learners can pause, repeat content, or move ahead as they feel confident. Follow-up check-ins and short refresh sessions help reinforce learning.

By offering modular and flexible pathways, AI training can reach more people, reduce drop-out rates, and support consistent learning across sectors.

3.5 AI training must be expandable

AI tools are being adopted quickly across organisations. They are rarely limited to one team or role. As a result, AI skills training needs to be able to scale across functions and levels of responsibility.

Training programmes should be designed for use in different work settings. This includes small community providers and SMEs, as well as large national employers. Training should work for varied roles, levels of digital experience, and sector needs.

One of the case studies in this report (Airbus) shows how scalable training can be applied across functions and locations.

Effective scaling relied on a clear core offer. Shared foundations help ensure consistent understanding and standards, while allowing flexibility where needed. Role-based modules and use-case libraries can allow the same training to be used across teams, with changes only where needed.

As AI tools become more widely used, training demand will increase. Training that cannot scale risks leaving parts of the workforce unsupported. This can lead to uneven skills across the workforce, poor practice, and increased organisational risk.

Scalable and expandable training helps organisations respond to growth in AI use without repeatedly redesigning training programmes. It supports coordinated rollout across teams and helps maintain shared standards as AI applications become more widely used.

3.6 AI training must be sustainable

AI training needs to extend beyond a single session. AI tools and applications are changing quickly, so training must be easy to update rather than fixed. Sustainable AI training needs to:

• be designed to evolve over time
• equip individuals with skills that remain relevant and adaptable as AI develops

One of the case studies in this programme (Roche) shows how ongoing training, reinforcement and clear governance can support AI capability over time.

Many organisations began by using generative AI tools that support tasks such as writing, summarising, or drafting content. AI is now moving towards more agentic forms, where tools can influence decisions, suggest actions, link tasks, and work with less direct input. Employers also report ongoing challenges in developing technical AI skills (67%), alongside gaps in responsible AI (32%) and non-technical skills (10%). As AI use changes, the skills people need will also change.

Training should be designed so it can be updated easily. Core content should be reviewed regularly to reflect new uses of AI. Revisiting training after 3 and 6 months can help staff reflect on how they are using AI and make changes to their way of working.

Sustainability also depends on understanding whether training is effective. Simple review methods, such as short surveys or team check-ins, can help employers assess:

• confidence in using AI
• quality and consistency of use
• impacts on judgement and decision-making

Responsible use must be built in from the start. This includes confidentiality, consent, data protection, transparency, and human oversight. These should be treated as part of everyday practice, not optional or advanced topics.

By combining regular updates, follow-up review, and clear expectations for safe use, AI training can remain relevant and trusted as tools and ways of working continue to change.

4. Common pitfalls to avoid

Evidence from the survey shows that AI training can deliver clear benefits. Over half of organisations (50%) reported increased productivity and efficiency, a further 26% reported that training enabled new responsibilities or projects, and 12% reported improved confidence with digital tools. This suggests that AI skills can improve productivity and expand the scope of work.

However, employers also report lack of representation in training provision (42%) and limited digital skills within the workforce (28%) as key constraints, reinforcing the need for more inclusive and accessible training models.

Workshop experts highlighted that many organisations struggle to design and deliver effective AI training. The following pitfalls are commonly observed across organisations and sectors.

1. Training is not linked to real work - Training is too general or tool-focused, making it hard for employees to apply in their roles
2. Training is not inclusive or adapted to different needs - A single approach is used for all employees, without considering confidence, access, experience, or accessibility needs 
3. Employees lack time, support, and access to learn - Training assumes staff have time, devices, and stable connectivity, which is not always the case
4. Existing skills and informal AI use are not recognised - Some employees are already using AI, but this is not built into structured learning or progression
5. Training is poorly communicated or feels irrelevant - Technical language and lack of representation reduce engagement, especially for non-technical roles
6. Training is too fast, too technical, or assumes prior knowledge - This discourages participation, particularly for those with less digital experience
7. Training is not tailored to sectors or job roles - Generic content does not reflect real tasks, especially in regulated or safety-critical environments
8. Training focuses on tools rather than transferable skills - Over-reliance on specific tools limits long-term capability as technologies change 
9. Training quality is uneven and difficult to assess - Fragmented provision makes it hard to identify trusted, relevant training
10. Organisations lack data readiness and infrastructure - AI training is less effective where basic data systems are weak or disconnected
11. Leadership support for AI training is weak or inconsistent - Training is less effective when leaders do not prioritise it, link it to organisational goals, or model its use
12. There is no clear governance or strategic direction for AI use - Unclear rules, responsibilities, and goals limit adoption and reduce confidence
13. Impact is not measured and training is not sustained - Without clear outcomes, training is deprioritised, and time, cost, and capacity constraints limit scale

5. How to use the tools together 

The AI Skills Tools Package, introduced in the AI Skills for the UK Workforce report includes 3 tools:

• the AI Skills Adoption Pathway
• the AI Skills Framework
• the Employer AI Adoption Checklist

The Tools Package supports planning and decision-making, while PRIMES guides the design and delivery of effective training in practice. Together with the PRIMES framework presented in this guide, these provide a practical and structured starting point for employers.

A typical approach would be:

Identify your starting point

Use the AI Skills Adoption Pathway to understand your current stage of AI use:

1. Awareness - Understand what AI is and why it matters
2. Exploration - Identify potential applications in your context
3. Assessment - Assess existing AI skills gaps and needs
4. Experimentation - Trial small-scale tools
5. Reflection - Learn from early experiences
6. Upskilling - Build staff AI capability and confidence
7. Integration - Embed successful tools into workflows
8. Strategy - Align AI use with your organisation’s goals
9. Scaling - Expand adoption across teams and services

Survey findings provide important context on where most organisations are starting from. AI adoption is still concentrated in early stages, particularly awareness (21%) and exploration (19%), with far fewer organisations reaching more advanced stages such as integration (7%), strategy (4%), or scaling (1%).

Define skills needs

Use the AI Skills Framework to identify the technical, non-technical, and responsible AI skills required across roles.

Plan actions

Use the Employer AI Adoption Checklist to prioritise actions and next steps.

Design training using PRIMES

Apply the PRIMES principles to ensure training is practical, reachable, integrated, modular, expandable, and sustainable.

PRIMES acts as a quality layer, ensuring that training designed at each stage is effective, impactful, and sustainable in practice.

Together, these steps provide a structured way to move from early experimentation to more consistent, effective, and responsible use of AI across the organisation. Progress does not need to be fast. Small, practical steps can help build confidence, improve skills, and support steady progress.

6. Cross-sector training patterns

Across all 10 sectors, the use of AI is increasing. Generative AI is already used in everyday tasks such as writing, reporting, and communication. More advanced and agentic forms of AI are emerging, but their use is still at an early stage in most organisations.

A consistent finding is a gap between AI use and structured training. In many cases, employees develop AI skills through trial and error, peer support, or built-in tools. This means there is often no shared understanding of what safe and effective AI use looks like in practice. Training needs vary depending on how work is organised.

In sectors such as construction, social care, and the creative industries, work is often fragmented. Many workers are in small organisations, freelance roles, or subcontracted work. In these settings, training needs to be simple, flexible, and closely linked to everyday tasks.

In sectors such as health, life sciences, defence, and financial services, work is more regulated. In these environments, training needs to be structured and aligned with rules, governance, and professional standards. There is a strong focus on safe use, oversight, and accountability.

In sectors such as manufacturing and clean energy, training needs to support real-world operations. This includes interpreting AI outputs, maintaining quality, and ensuring safety in day-to-day work.

In digital and professional sectors, job roles are changing quickly. There is a growing need for skills in oversight, verification, and working effectively alongside AI systems.

Despite these differences, there are clear shared priorities across all sectors. AI training needs to:

• be linked to real tasks and workflows
• build confidence and judgement, not just technical skills
• include responsible and safe use of AI
• be accessible and flexible in how it is delivered
• support different roles and levels of experience

Evidence from case studies shows that organisations that move from informal use to structured training achieve more consistent results. They report improvements in productivity, confidence, and quality of use.

For a full overview of sector-specific challenges, AI use, and training needs, see the research evidence, analysis and methodology document. For detailed examples of how organisations are developing AI capability in practice, see the supporting case studies document.

7. Where to see case studies working

The case studies in this programme show how organisations are already developing AI skills in practice. They cover a range of sectors, organisation sizes, and training approaches.

These examples are not intended as a complete list. Instead, they show what effective AI upskilling looks like in real work settings, and how organisations move from early use to more structured capability building.

Each case study follows a consistent structure to support comparison and practical use. This includes:

• the challenge the organisation faced
• the intervention introduced
• the nature of the journey
• the observed impact
• the key outcomes and lessons

Together, these cases show not only what was done, but what worked and why. They provide clear insight into how different approaches to AI upskilling deliver value, and the conditions that support effective and sustained capability building.

7.1 Case studies grouped by model

The case studies are grouped into 4 models reflecting different organisational contexts, capability needs, and approaches to AI upskilling. Each group highlights how organisations respond to distinct challenges and build capability in practice.

Large enterprise, regulated environments

This group includes large, complex organisations operating in highly regulated sectors, where AI adoption must align with governance, compliance, and long-term workforce transformation. These organisations focus on building AI capability at scale, embedding learning within organisational systems, and ensuring responsible and controlled use of AI.

The case studies in this group are:

• Airbus - building AI capability through long-term workforce transformation, with structured learning pathways, competency frameworks, and organisation-wide integration
• KPMG - embedding AI upskilling as an organisational capability through leadership sponsorship, workflow-integrated learning, and applied use cases
• Roche - delivering enterprise-wide AI capability through mandatory training, governance frameworks, and responsible AI practices

These examples show how AI capability can be developed systematically across large workforces, with strong alignment to strategy, governance, and organisational performance.

SMEs, creative industries, construction and consultancies

This group includes smaller organisations and consultancies operating in dynamic, resource-constrained environments. AI upskilling is typically more agile, practice-based, and closely tied to immediate business needs and workflows.

The case studies in this group are:

• Vertis Media - a creative SME building in-house AI capability through internal training, shared learning structures, and workflow integration
• Congregation - a learning and transformation consultancy supporting organisations through applied, behaviour-led AI upskilling and organisational alignment
• Cast Consultancy - a construction consultancy using a use-case-led approach, focusing on specific workflows (for example Gateway 2 submissions) to build practical and responsible AI capability

These examples demonstrate how organisations build AI capability through experimentation, applied learning, and direct integration into business processes, often without large-scale formal infrastructure.

Platforms, training providers, and activation models

This group includes organisations that provide structured learning environments, frameworks, and activation models to support AI capability development across different sectors and workforce groups.

The case studies in this group are:

• LinkedIn Learning - supporting organisational AI upskilling through structured learning frameworks, capability levels, and personalised learning pathways
• 100 School - delivering challenge-based, behaviour-focused AI training programmes designed to drive adoption, habit formation, and real-world application

These examples show how external platforms and providers can support organisations in moving from informal experimentation to structured, scalable AI capability development.

Public sector and inclusion-focused models

This group includes organisations focused on public service delivery and digital inclusion, where AI upskilling must address accessibility, trust, and the needs of diverse and underserved populations.

The case studies in this group are:

• NHS Primary Care - supporting AI upskilling through practice-level adoption, with a focus on improving efficiency and patient interaction within clinical workflows
• Good Things Foundation - addressing AI literacy gaps through inclusive, community-based learning models designed for digitally excluded groups

These examples highlight how AI capability can be developed in ways that support service improvement, inclusion, and equitable access to technology.

For a full overview of these examples, see the supporting case studies report.