The CBI on AI: Skills, Jobs and the £47bn Opportunity

What’s the one difference that means some UK businesses get significantly more from their AI investments?

In this conversation, Richard Nicholas is joined by Shivani Gupta, Senior Policy Adviser for AI and the Digital Economy at the Confederation of British Industry (CBI). With daily access to Whitehall and the businesses the CBI represents, she brings the findings of the CBI's AI skills report - and a reality check the headlines rarely give you.

 The biggest blockers aren't technical but rather human: poor data quality, governance gaps, and a cultural shift most boards haven't made yet.

An insight that lands with particular force. AI won't take your job, but it will take the tasks that teach you how to do it. That has real consequences for early careers, onboarding, and how organisations build expertise when routine work disappears.

 With only 39% of UK businesses use AI in a meaningful way, Shivani's advice to business leaders is direct. Don't wait for perfect conditions - they're not coming. Start small, involve your people early, and invest in skills alongside the technology.

 Essential listening for business leaders and anyone navigating AI strategy, workforce transformation, and responsible adoption.

Transcript

Welcome And Meet The CBI

Richard 0:09

Okay, so welcome everyone to this podcast, m y name's Richard Nicholas, and I'm the host of this session. Today I'm here with Shivani Gubter, who is senior policy advisor for the CBI. The CBI having just carried out an AI skills report, which we'll learn a little bit more about later. So Shivani, good to meet you. I wonder if you could give us a bit of an introduction about yourself and the CBI and your role within it if that works.

Shivani 0:41

Yep. Thanks, Richard. Thanks for having me. It's great to be here. On the CBI, yes, of course. So the CBI, the Confederation of British Industry is essentially the voice of business in the UK. We act as a catalyst between government and industry, and we are the voice of about 170,000 businesses, which are of all sizes and and across sectors of the UK economy. And our corporate members alone employ about 2.3 million workers. So when we speak, we speak for a significant slice of the UK economy. And I think the way we describe ourselves really resonates with what where we are at the moment. It's an unstable world politically, technologically, as well as economically and business voices are important more than ever before. They have such an important role to play when it comes to growth, jobs, investment, productivity. And our job at the CBI is to make sure that that voice is heard by government and it reflects in better policies. I lead a work on AI and the digital economy. I sit within the CBI's tech and innovation team. And my work involves covers everything AI. I work on a daily basis with government on different aspects of AI policy. Alongside I also work with businesses listening to what's needed in real, in ground reality, how they're experiencing things. So, it's a varied role, and the way AI is advancing, there are no two weeks that actually look the same, but it's a very important role. I think, I enjoy my work.

Richard 2:22

No, it sounds it, and having that sort of overview of what's going on in business sounds you know particularly useful in terms of AI, you know, different sectors at different stages. You'll be seeing, you'll be getting that sort of overview. And presumably that's what led to the AI skills report that the CBI published. Now be interested to know sort of what led the CBI to focus on AI skills particularly, and why has this become such an important issue for businesses right now?

Why AI skill are important?

Shivani 2:55

Yes, certainly. I think we started doing this work by just because we were listening to members, we were listening to businesses. The policy conversation around AI was moving incredibly fast. Government was coming up with, or we are still coming up with strategies, initiatives, action plans, left, right, and center. And while we appreciate the ambition there, we felt that there's something missing in terms of delivery, that there's a certain gap there and when you unpack that, that really comes to a missing employer voice. Because the businesses were asking us how are other businesses approaching AI workforce transformation? How do they begin, how do they tackle risk? And there were no clear answers to that. So this led us to bring together what we called a task and finish group on AI skills. It had 11 business leaders from across sectors, including life sciences, including manufacturing, logistics, universities. So it was a very diverse group. But interestingly, we didn't have a single tech member, and that was a deliberate choice because we wanted a demand side perspective and not a supply side where a large tech vendor would tell you how businesses should be doing that. So, yeah, so that is how we started this work, and it kind of expanded into a report that we published early this year. And the second reason was that the urgency was real. It is about 39% of UK businesses actually use AI in a meaningful way. The rest are not, they are experimenting with it, but it's not leading to any return on investment, and the gap between early movers and everyone else was widening. And, if you look at what is at stake, the government recent government analysis says that AI could add about 47 billion to UK economy every year. You start to understand why getting the skills pieces right, because that price can only materialize if the workforce is ready for it. So these are the two reasons which led us to start this work, and it has since become quite a broad work stream for us.

Richard 5:10

Yes, I can well imagine. 39% is I don't know, that seems relatively low when you see all of the comments about AI and all of the news stories for such a long time, it feels like. So it's interesting to have that figure, I suppose. And I'm no doubt that will grow over time. But what are businesses actually telling you about how AI is changing work? Are we talking about job losses or is it a bit more nuanced than that?

AI At Work Beyond Job Loss

Shivani 5:44

Yeah, I think it's a bit more complex and nuanced than that. While the mainstream media and the public narration talks a lot about job losses and large-scale headcount reduction, we are supposedly, we are all losing our jobs in the next 10 years but I think that doesn't accurately capture the real picture because when we speak with businesses, most of them are not doing reducing headcounts. They're actually allowing some roles to kind of phase out in the long run while they are restructuring existing roles. For instance, there's a like a pharmaceutical business who is in our membership. They are running a transformation program that they call work deconstruction. Interestingly, they are looking at each jobs as a series of tasks and then mapping each task to see whether it would best be done by AI or would a human involvement be required. In fact, the ONS data also shows that only 5% of UK businesses have actually reduced headcount owing to AI, rest have not, which was surprising given the narration that we hear about how AI is taking over our jobs. And in my experience, even with businesses that have started realizing the return on investment, for instance, there's a logistics company which have started to employ AI driven robots within their warehouses. They told me that they have doubled their productivity owing to that and have reduced vehicle damage by 40%. So that's a huge productivity increase but this has not led to a headcount reduction. In fact, they have retrained the warehouse workers to actually manage those AI powered robots. So it's they're looking, it should, it's about more about augmentation where you're working alongside AI rather than competing against it, but having said that, I do not deny that there are certain roles that are at risk, specifically entry-level roles, and the ones that are repetitive in nature, they are at risk and people are losing that job, those roles. So it is important for us to kind of upskill ourselves so that we we still remain relevant as as this technology advances further. So yeah, I think it's a mix of both, but certainly it's not all only about job losses, it's a lot more than that.

Richard 8:20

Well, that's good to hear. Reassuring words across the industry. So only 5% actually resulting in five-finding job losses. Again, not the picture that you would get from sort of the regular feed on LinkedIn at least. That's one surprising thing.

The Hidden Risk For Juniors

Richard 8:48

I suppose what else did you find out that was sort of surprising or unexpected from all the conversations that you had in putting the report together?

Shivani 8:57

Yeah, one thing, one insight that has actually stayed with me, and I will quote because I think it is powerful, is that AI will not take our jobs, but it will take the task that teach us how to do that job. So when you think of that, it is actually so true because things like drafting reports or scheduling or collecting data, things that junior employees used to do to kind of learn their work, to understand how the organization works, to build opinions on certain things and to make risk in relatively low-stake environment, they're all being overtaken by AI. So while we talk about mass redundancies, we talk about reducing headcounts, we are forgetting that there is a whole generation of workers coming up who do not have the domain knowledge that the older generation have because they aren't doing that work. So we need to rethink how do we think about formative learning because we're not, this is not something that has still come up that problem that we have already started to face, because this is something that I feel that we will see in the next five to ten years when those people are reached to the leadership level. So it is important to think about early career pathways. It is and it is important to think how you are onboarding your employees currently within the organisation. One of members that I engaged with told me that their early career pathways has changed from a junior employee doing such kind of data collection and drafting and scheduling to actually becoming a manager from his very first job. And what does he manage? He does not manage humans, he is managing those AI agents and AI robots that exist in the work stream. So it's like something that you were expected to do at a much later stage. You are now expected to do at a junior level, but with a very distinct flavor to it because you're not managing humans anymore. So I feel that's one thing that we need to think, and it was surprising when it actually came out that how businesses are actually seeing that. And the other thing I would say that kind of surprised me was the reasons for resisting AI. So we often hear about the hostility that employees have around AI being implemented or adopted within a company or an organization. But when we spoke with some employees, it seemed that the resistance comes from change fatigue as well as because they do not understand how AI is going to affect their role. They see it as a threat because they do not know whether their skills is going to be relevant anymore or how is it going to change their jobs. And so for instance, and it is so it I feel it's important for businesses to actually have an open conversation and with employees as to what is, what are the changes that they are expecting and take them along the journey rather than just impose it upon them. A gain, and I'll give you an example which kind of stayed with me. One of the our members formed what they call a citizen building forum, which is essentially a group of people, employees coming together, irrespective of what team they work in, irrespective of whether they are technicians or non-technicians, they came together to build tools to solve problems that they face on in their day-to-day life. And this led to hundreds of tools that they built, which led to a massive increase in productivity. So I think there's something to learn here. If you take if you see it as a democratic process from bottom up, employees will not resist it, but see it as a companion rather than something they are competing against.

Richard 12:58

Yes, I think I've heard that in a few places. So a sort of bottom-up adoption, getting people to actually create the tools themselves and share the tools, like you're saying, in order to get that adoption so they'll actually feel like they're involved in the process of actually taking on the change as opposed to having to just adapt to it. Okay. Yeah, yeah, yeah. That's interesting. And are there any other barriers holding businesses back in your view, or is that the main thing, the sort of reluctance problem?

Shivani 13:42

Yeah, yes, skills is of course an important barrier, but other than skills, I feel that data and governance is also something that members can consistently tell us is a barrier to adoption. And I feel that data and governance are the

Data Governance And Board Ownership

Shivani 13:50

unglamorous but critical foundation for AI adoption, so you might have the best tools in the world, best AI tools in the world b ut if your data is of poor quality, it is not managed well, the results you get out of your AI model is not going to be reliable. So it is important for to have a good quality, good well-managed data. And while most businesses in the UK have already started experimenting with AI, if not scaling, and those who have not are intending to do it, very few actually have internal data expertise and governance infrastructure. In fact, there's also a big part that the even the government can play here because there is certainly there are some regulatory uncertainties as well. Businesses working across jurisdictions, for example, in finance, global supply chains, often have to navigate a very complex and sometimes contradictory compliance expectations. So it is I feel that these are the two things that it is important for businesses as well as the government to kind of look into. The third thing I see, I feel is a cultural dimension. In most businesses, I see AI governance sits in their IT team and not in the board. But the way the technology is expanding, it has started to affect core business processes, how decisions are made, how you how do you serve your customers, how do you manage this. So these are not just IT questions, they are board level questions, and that shift needs to happen within businesses. So I think beyond skills, data and governance is something that is extremely important in terms of AI adoption. Without getting that right, we cannot get we cannot do adoption responsibly.

Richard 15:55

Yes, completely. No, I can, I completely agree. Actually, the idea of AI still being in a, in the IT team feels feels something of a few years ago. Every board director needs to use kind of skill care and diligence. The board needs to be aware of what's happening, it's a strategic decision to have.

Shivani 16:14

Definitely and I feel that leadership buy-in is also very important because, when I spoke about resistance, I feel that when you see your leaders actually using it and encouraging you, it's has a psychological effect on you to actually do it. So visible leadership buy-in is very important in when it comes to AI adoption and deployment with any business.

Richard 16:49

Excellent. No, completely. And so on the back of the work that you've done, so you've identified various things that need to be done, this upskilling, the getting the visible leaders effectively, getting AI champions effectively in the business. What's the CBI actually doing to help businesses make progress on AI adoption?

CBI Programs To Boost Adoption

Shivani 17:11

Yeah, quite a few things actually. So the report itself, the AI skills report that we published, was aimed to fill a gap. We wanted to provide a demand side perspective on how leaders are approaching workforce transformation. And it had a good impact. We were the recent announced at the Department of Science, Innovation and Technology, they invited us to be a strategic partner in the AI Skills Boost initiative. So the ambition there is to upskill 10 million workers by 2030. This is supported through an AI skills hub website, which has hundreds of courses. They are free of course, they are provided by the tech partners that DCT has. And you can just sign up and do the whichever course you feel is relevant for you. So the CBI's role here is to make sure that business perspective would inform delivery and to make sure that it reaches the right people, the harder-to-reach group, the catalyst roles, the SMEs. So to support that work, what we are launching is the AI B2B peer learning series. These will be delivered regionally but with a national impact. The idea is that because business our members told us during our engagement that they do not necessarily need more information on AI. What they would really value is gaining real insights, sharing real examples and practical approaches on how different businesses with facing similar challenges and navigating these issues. With so our aim is to provide a space and a structure for these conversations to happen at scale. So we've developed a few modules around which these conversations could kind of revolve around these include leadership and management, risk and governance, workforce transformation, and scaling use cases. So there's a bit there for everybody, I feel. And other than that, we are continuing our work with the government. We are doing some campaigning around some of the recommendations that came out in the AI skills boost in the AI skills report, including advocating for an AI literacy standard as a baseline skill that businesses can reach for. We are also speaking with the businesses on greater regulatory clarity. The CBI is currently form developing a policy position paper on AI regulations and governance. It's in the final draft stage, and we are just doing a final round of member engagement before it is published. So there are a couple of things that we are involved in. Potentially, we are also looking to do an AI adoption task force as well, which would cover, which would be a more broader, which would cover a more broader area around AI regulations, governance, role of UK in the international world, B7, B20, what the UK could do there. So there are a couple of things that we are trying out, but are the main goal is to help in responsible adoption across the economy. So everything that we are doing with it's with that aim in mind.

Richard 20:30

Excellent. No, and worthy aims as well, something goes very much to the focus of our business, which is that AI adoption, doing it responsibly. So if you were speaking to a business leader, if a business leader listened to this podcast who wants to get serious about AI adoption, what would you want them to take away from this conversation? What should they do next?

Shivani 21:02

Yes, a couple of things actually.

What Leaders Should Do Next

Shivani 21:04

The first I would say is that don't wait for perfect conditions because they aren't coming. The businesses that are the most further ahead in this journey are the ones who started with a small pilot. They learned fast and scaled what work. The approach that came very consistently from this work was that the best approach was to pilot test and scale, not big bang rollouts, not waiting until you've got everything figured. Just start, learn, and build from where you are. The second thing is to involve people from the very beginning, not once the technology is already decided, not as a consultation exercise after the fact, but actually ask your employees that which part of their roles could most be supported through AI and where would they value being in control and implement your AI accordingly. Thirdly, I would say that invest in skills alongside technology. A mistake that we often see businesses doing is that they first bring in the technology, invest in it, and then this people don't know how to use it, or they just they refuse to use it because they see it as a danger. So it is important for you to for businesses actually to invest in their people alongside the technology. Lastly I would say that it is important to be ambitious as well because I see a lot of businesses view AI as a cost-cutting tool, which is fine, it works that way to an extent, but that is not the most transformational path. When the the real dividend of AI comes when you think how it can change business models, how value can be created, and how it can be used in changing the ways of working overall. So I know that this involves a lot more commitment than looking at it as a cost-cutting tool, as a cost-out tool, but I feel that that's where the real dividend is, that's where the real price would materialize. So, and that is what we are trying to do through the skills program that we are running, through regulatory clarity campaign that we are running, because we want AI to be AI has the potential to offer incredible growth to the UK. And we feel that, but it should be used in a meaningful way that it can actually translate and actually release the potential that it has.

Richard 23:41

Excellent. Superb. So Shivani, yes, investing in skills, investing in your people and and being ambitious, they all sound like good plans for any business leader, anyone who would like to get in touch with you or the or the CBI more generally, presumably they should look on the website. I'll put details of those on the notes to this to this podcast and yes, any final thoughts before we close

Final Advice And How To Connect

Shivani 24:20

Yeah, thank you so much for having me, it was a very good session. I but I think I'm very open to having further conversation with any business leaders who would want if they're keen to being involved in the B2B series or they are keen to be just to hear what more we are doing in the space. Very happy to have a have conversation. And I think the final thought would be do, not be scared of AI, look at it as a companion, don't see it as a competition. It it and it would actually help in increasing productivity, it would help add value to the work that you do. So don't be run by business by public narrations and try and upskill yourself so that you're relevant in the AI world.

Richard 25:01

Fantastic. Excellent, Shivani. Thank you very much. Absolute pleasure to speak to you today, and hopefully we shall we shall speak again soon.

Shivani 25:12

Thank you. Thank you for having me, same here. Bye.