EP23: Reviewing the AI Opportunities Action Plan: What do Businesses Need to Succeed?

AI Adoption, Talent and Investment: Building the UK’s Next Generation of AI Businesses

As the UK moves from AI ambition towards implementation, questions around adoption, talent and investment are becoming increasingly important. At a recent UKAI discussion, investors and founders explored what businesses need to do differently to make AI useful, how government can support adoption, and whether the UK has the talent and capital required to build globally competitive AI companies.

Opening the Floodgates to Grassroots AI Adoption

For Zoe Qin, Vice President of Dawn Capital and former software developer, one of the biggest opportunities is to encourage employees to experiment with AI themselves.

Rather than relying entirely on top-down implementation, businesses can create environments where teams identify small problems in their everyday work and explore how AI might solve them. She pointed to examples from legal and medical environments, where professionals have increasingly become power users of AI tools.

Internal hackathons and “hack nights” can provide a practical way to encourage this experimentation. Giving employees dedicated time, tools and support helps them move beyond recognising that something could be automated to actually building a solution.

The approach also recognises that AI adoption is rarely one-size-fits-all. The final stage of implementation often requires employees to customise tools around their particular workflows.

Closing the Gap Between AI Hype and Reality

For Hussein, founder and CEO of Quench.ai and previously co-founder of AI identity-verification company Onfido, one of the biggest barriers to adoption is the gap between what businesses are being promised and what AI currently delivers.

The continued narrative around imminent mass automation can create unrealistic expectations. While major technology companies frequently emphasise how rapidly AI capabilities are advancing, businesses on the ground are often still using AI in relatively basic ways.

For many organisations, the most successful applications are currently embedded within tools they already use, rather than entirely new AI systems. This disconnect between hype and practical value can make businesses uncertain about where AI genuinely fits.

Treating AI More Like a Resource Than Software

Another challenge is that AI does not always behave like traditional enterprise software. Traditional software is generally purchased to perform a defined task, installed and maintained. Agentic AI, by contrast, needs to be trained, monitored and continuously improved.

Hussein argued that businesses should therefore think about AI less as a product that is simply procured and more as a resource that needs to be managed over time.

This also changes the way organisations should approach adoption. Rather than asking, “How can we use AI in our company?”, businesses should start with their core competencies and value propositions and work backwards from the problem they actually need to solve.

The objective should be to use the least amount of AI necessary to improve the desired outcome, rather than introducing AI simply because the technology is available.

Turning Experimentation Into Adoption

For employees who already know there are parts of their work they could improve, the biggest obstacle can simply be finding the time to experiment.

Zoe described hackathons as a way of creating that space. Participants can bring small frustrations or repetitive tasks and spend concentrated time developing solutions, often using no-code tools that make experimentation accessible beyond traditional technical teams.

The model can also bring employees together with AI vendors and technical experts, creating an environment where businesses can move from curiosity to practical experimentation. This grassroots approach may ultimately be more effective than expecting leadership teams to design every AI application from the top down.

Why Government Needs to Experiment Too

Government faces a different challenge. The consequences of a failed AI implementation can be significantly greater than in a private business, particularly when public services or sensitive information are involved.

Hussein argued that this makes a slower, more cautious approach understandable, but pointed to regulatory sandboxes as a way of balancing experimentation with protection.

The UK’s experience with fintech provides a useful precedent. Rather than preventing innovation altogether, regulatory sandboxes can allow new technologies to be tested in controlled environments, helping regulators, procurement teams and government departments learn what works before scaling it.

Good regulation, he argued, should not necessarily slow adoption. If designed effectively, it can actually give organisations greater confidence to innovate.

Making AI Benefits More Visible

The AI industry also has a role in helping government understand what successful adoption looks like.

Rather than focusing exclusively on technical capabilities, businesses and investors can communicate through tangible use cases. Hussein suggested that AI will become easier for the public and policymakers to understand when they can see direct benefits in areas such as healthcare, education and waiting times.

This reflects a broader challenge for AI adoption: people need to understand what the technology changes in their own lives. Stories can make the technology more tangible, helping policymakers and the public move beyond abstract discussions about AI towards a clearer understanding of its practical value.

Talent: Attracting the People the UK Needs

Talent remains one of the central challenges for the UK AI ecosystem.

Zoe highlighted the importance of attracting international talent, drawing on her own experience of working in the UK on a skilled worker visa. She argued that visa processes and the wider environment for international workers can influence whether highly skilled people choose to build their careers in the UK.

The UK’s Global Talent visa was highlighted as a positive route, but attracting talent is only part of the challenge. The country also needs to make its own technology sector more visible to people entering the workforce.

Startups can struggle to compete with major technology companies for attention among students. Zoe argued that companies need to engage with universities and younger talent much earlier, making the range of opportunities in the UK’s startup ecosystem more visible.

Hackathons, showcases and founder events can provide a bridge between students and emerging companies, allowing young people to discover opportunities beyond the most recognisable global technology brands.

Developing Talent, Not Just Importing It

For Hussein, the UK’s talent challenge goes deeper than immigration. The country has strong universities, scaleups and an established technology ecosystem, but the long-term supply of skilled workers depends on the education system.

He distinguished between attracting talent and developing talent. While international recruitment can help companies meet immediate needs, the UK also needs to prepare young people entering the workforce for an increasingly AI-enabled economy.

This requires thinking over decades rather than electoral or five-year cycles.

Hussein argued that startups should not be expected to solve the entire skills problem themselves. Young companies are often operating at the frontier of innovation and need people who already have a strong foundation of skills. Building that foundation therefore has to begin much earlier.

Rethinking Education for an AI-First World

AI may also require a different approach to education itself.

Hussein described three broad forms of intelligence: knowledge and summarisation, procedural understanding of how to perform tasks, and a more contextual form of intelligence involving perspective, relationships and reading between the lines.

As AI becomes increasingly capable of handling knowledge-based tasks, he argued that education should place greater emphasis on creativity, social interaction, play and other forms of human development.

The objective is not simply to teach young people how to use AI, but to develop the capabilities that complement it.

The UK’s Investment Landscape

Despite concerns about the UK’s ability to compete globally, both speakers argued that the country does not necessarily suffer from a shortage of capital.

Hussein suggested that the deeper issue is human capital: having enough founders and senior operators capable of turning investment into successful businesses.

An excess of funding without enough strong companies can create a cycle in which too many startups are funded without developing sustainable businesses. The goal should therefore not simply be more startups, but more companies that successfully scale, become profitable and generate significant exits.

Zoe similarly argued that the UK’s investment environment is more competitive than the common perception that entrepreneurs have to go to Silicon Valley to raise serious funding.

She noted that AI accounted for around 30% of UK venture funding in the first half of the year discussed at the event, representing a record share according to the figures she referenced.

From AI Applications to Fundamental Innovation

Investment is also beginning to move beyond companies simply adding an AI feature to an existing product.

While application-layer businesses remain an important area of investment, Zoe highlighted growing interest in more fundamental technological challenges, including scientific innovation such as optical computing.

This reflects a broader maturation of the AI market. Investors are increasingly considering not just how AI can be incorporated into existing software, but what new infrastructure, scientific breakthroughs and business models it can enable.

At the same time, AI-enabled services are creating opportunities to make traditionally labour-intensive businesses more scalable by using AI as an additional resource.

What Makes a Successful AI Founder?

For Hussein, successful founders need to combine several seemingly contradictory qualities.

They need to be able to take a high-level view of a business while understanding the details of the problem they are solving. They need creativity alongside analytical thinking, as well as the determination to pursue a difficult problem over a long period.

But ambition also needs to be balanced with humility. The strongest founders, he suggested, are continually curious about their subject. Their interest extends beyond their working hours because they are genuinely invested in understanding and solving the problem at hand.

Moving Beyond the Fear of AI Replacing Jobs

The discussion also turned to people who have little interest in technology and may feel excluded by the pace of AI development.

Hussein argued that the dominant narrative needs to change. The future is not necessarily one where humans are simply replaced by AI. Instead, people who use AI effectively may increasingly compete with those who do not.

That does not mean everyone needs to become an AI expert. Just as the internet became embedded into everyday life without requiring everyone to become technically proficient, AI will become a background capability across many jobs.

The challenge is ensuring that people are given the opportunity to develop enough AI literacy to participate in that changing economy.

Making AI a Personal Experience for Policymakers

One of Zoe’s suggestions for government was surprisingly simple: policymakers should consider how they personally use AI.

Understanding AI through everyday experiences can make the technology less abstract and help leaders recognise the wider range of applications available.

From there, the policy challenge becomes creating an environment where people can turn those experiences into businesses — making it easier to start companies, experiment and build new products in the UK.

Building the Next Generation of UK AI Companies

The discussion ultimately returned to a question of what the UK needs to prioritise.

The country has capital, universities, entrepreneurs and an established technology ecosystem. But sustaining that advantage will require more than increasing investment alone. It means creating an environment where people can experiment with AI, developing talent from an early age, attracting skilled workers from around the world and making the benefits of AI visible through real-world stories.

For businesses, the message is equally practical: start with the problem, not the technology. Encourage employees to experiment, give them the tools to build and focus on measurable outcomes.

The opportunity for the UK is not simply to adopt AI faster. It is to create the conditions for the next generation of AI companies, founders and workers to build what comes next.