AI HackathonsConsultancyAbout

    Blog · AI Implementation

    What will it take for UK businesses to get a return on AI adoption?

    · By Day Seven

    Short answer: New modelling for the British Chambers of Commerce says wider AI adoption could make the UK economy more than 10% larger by 2031, but only if businesses actually put AI to work. Recent UK workforce research suggests that the return depends on skills, training and clear ownership of AI capability, not just on the tools a business buys.

    What the BCC published

    On 4 November 2026 the British Chambers of Commerce (BCC) published new analysis on AI adoption. The modelling looked at two adoption scenarios, which the BCC says are based on research by Anthropic:

    • Higher adoption. If AI boosts productivity by 57% across 12% of work tasks by 2030, UK GDP would be more than 10% higher by the end of 2031 than it would otherwise be.
    • Moderate adoption. If AI boosts productivity by 35% across 4% of work tasks by 2030, UK GDP would be around 2% higher by the end of 2031.

    The gap between those two outcomes is the point. The BCC notes that the Budget the week before included over £700 million of AI-related investment, mainly in infrastructure and security. Its Director General, Shevaun Haviland, said business adoption is "the missing piece of the AI jigsaw" and called for an adoption strategy that "focuses on skills". The analysis comes alongside a BCC book of essays, produced with BDO, on how the UK can become an AI adoption superpower, and a BCC AI Adoption Summit in London on 8 December.

    Why adoption stalls inside organisations

    Modelling at the level of the whole economy can feel distant. Two recent pieces of UK workforce research show what slow adoption looks like inside a single business.

    Kaplan's research on AI capability. Kaplan's whitepaper, From AI to ROI: Who owns the problem?, is based on Censuswide surveys carried out in July 2026 of 1,000 UK employees, 600 HR and L&D decision-makers and 600 C-suite and director-level respondents, all in organisations of 50 or more people already using AI. As reported by Learning News, 37% of senior respondents said their organisation had missed expected returns from an AI tool or deployment in the past 12 months because of AI skills gaps. 54% said their AI investment had placed more emphasis on tools and technology than on workforce capability, against 20% who said the reverse.

    As covered by theHRD on 4 November, the same research found that 94% of HR and L&D leaders say they understand the capability their AI strategy needs, yet 36% lack the budget, the authority or both to act fully on it. A third of employees (33%) said they had been asked to use AI without sufficient training, and 45% of HR and L&D leaders said AI tools were being underused.

    There is also no agreement on who is responsible. 44% of C-suite respondents said IT or technology is the main owner of AI capability, with only 11% naming HR and 6% naming L&D. HR and L&D respondents were more likely to name HR (26%) or L&D (18%).

    O'Reilly's research on everyday use. In a survey of 2,000 UK employees published on 27 October, O'Reilly found that 39% of workers now turn to an AI chatbot first for work answers. Yet 35% said their organisation had provided no AI training or guidance in the past 12 months, and only 16% had received training specific to their role. Of those using AI-generated answers at work, 39% said they had acted on one that later proved inaccurate, incomplete or unsuitable, and only 24% said they always verify an AI answer before using it.

    These figures are what people reported, not independent tests of skill or verified losses. Even so, they point in the same direction. Many organisations have bought or allowed AI tools, but have not yet built the skills, habits and ownership to get value from them.

    Practical takeaways by role

    L&D leaders: Kaplan reports that 81% of employees trained in the past year rated their AI training as effective, but only 35% said structured training had reached a significant share of the workforce. The quickest win is usually reach, not a new course. Map who has had practical AI training, who has not, and which roles use AI most, then close the biggest gaps first.

    HR leaders: Make AI capability part of role expectations and development conversations, not a side project. Agree a simple standard for what "good use of AI" looks like in each role, and back it with a clear employee AI policy so people know what they can and cannot put into AI tools. Our page on AI training for HR teams covers where HR can start.

    IT leaders: Tool rollout and training should run together. If your team is responsible for the AI tools, agree with HR or L&D who owns the people side before the next rollout, and share usage data so training goes where the tools are underused. Build checking habits into the training, since O'Reilly's research suggests most staff do not always verify AI answers.

    Transformation leaders: Treat ownership as a decision to make, not something to discover later. Name one accountable owner for AI capability, give them a budget, and agree how you will measure progress. Our guide to measuring AI training ROI explains how to set a baseline and track time saved and quality before and after training.

    A simple checklist for getting a return on AI

    1. Name an owner. Decide who is accountable for AI skills across the business, and how IT, HR and L&D share the work.
    2. Fund people as well as tools. When you budget for an AI tool, budget for training the people who will use it.
    3. Train by role. Build sessions around the real tasks each team does, not general features.
    4. Teach checking. Make verifying AI output a standard step, with clear rules on when a human must review it.
    5. Measure and refresh. Set a baseline, check usage and results a few weeks later, and update training as the tools change.

    What this means for your team

    The BCC's modelling shows how much is at stake if UK businesses adopt AI well, and how little changes if they do not. The workforce research suggests the main barrier is not access to tools. It is that training, budget and ownership have not kept up. If your organisation has AI tools in place but patchy use, the next step is probably a clear owner and practical training for the people expected to use them.

    For more context, see our analysis of the government's AI adoption survey 2026 and our business AI adoption guide. Not sure where your team stands? Try our free AI readiness scorecard, read about our AI adoption and AI upskilling support, or browse our AI training for business, including AI training in Leeds.

    Talk to us

    If you would like help turning AI tools into real results for your teams, book a free 60-minute discovery call with Qasim.

    Frequently asked questions

    What did the British Chambers of Commerce say about AI adoption?

    On 4 November 2026 the BCC published modelling which found that, in a higher adoption scenario, UK GDP could be more than 10% higher by the end of 2031 than it would otherwise be. In a moderate scenario the figure was around 2%. The BCC called business adoption the missing piece of the UK's AI plans and said an adoption strategy should focus on skills.

    Why do some businesses not see a return from AI tools?

    In Kaplan's UK research, 37% of C-suite and director-level respondents said their organisation had failed to achieve expected returns from an AI tool or deployment in the past 12 months because of AI skills gaps. The same research found 45% of HR and L&D leaders said AI tools were underused by the teams expected to use them.

    Who should own AI training in an organisation?

    There is no single right answer, but someone needs to own it clearly. Kaplan's research found leaders disagree: 44% of C-suite respondents named IT or technology as the main owner of AI capability, while HR and L&D respondents were more likely to name HR or L&D. A shared plan, with a named owner and a budget, works better than leaving it unclear.

    Do employees need AI training if they already use AI?

    Yes. In O'Reilly's survey of 2,000 UK employees, 39% of those using AI-generated answers at work said they had acted on one that later proved inaccurate, incomplete or unsuitable, and only 24% said they always verify an AI answer before using it.

    Does AI training actually work?

    The research is encouraging. Kaplan reports that 81% of employees trained in the past 12 months described their AI training as effective, and 63% of HR and L&D leaders who had provided AI training reported measurable improvements in workforce capability. These are self-reported figures rather than independent tests.

    Related posts