In the UK, 42% of businesses’ AI projects failed in 2025, a 25% increase from 2024. Despite this, over nine-tenths of business leaders say they are ready to integrate AI into their business. So, where are business leaders going wrong, and what can be done to rectify this alarming number of failed projects? Too often, leaders have overly ambitious visions of what AI can achieve in the first one or two steps. AI is not a plug-and-play tool. Instead, need to take a more pragmatic approach when leveraging AI in order to deal with the challenges they face.
AI offers extraordinary business value, yet its success hinges on the human factor. Its development is far quicker than the human ability to make use of it. This is where taking a pragmatic approach will be crucial for business leaders to close this gap between technological development and human ability.
Set expectations from day one and balance AI hype with pragmatism
One of the biggest challenges facing over 85% of business leaders is the pressure they experience from executive teams to adopt AI at speed and scale. Many expect instant transformation, but these unrealistic expectations can just paralyse AI’s ability to be effective. AI’s impact won’t happen overnight, and here’s when leaders need to set expectations from the get-go.
When integrating AI, leaders must look to prioritise manageable bitesize steps rather than trying to push through an all encompassing adoption. Steps must be formed from human, technological, regulatory, and security perspectives in order to drive productivity, quality, output, and cost effectiveness, and ensure security and regulatory compliance. Think AI stock management in the retail industry or predictive maintenance in the manufacturing industry. These are good examples and provide a workable pragmatic approach to achieving AI success.
At this stage, do’s and don’ts have been learnt, impact has been witnessed, and workforces have experience of working with the tools. From here, leaders can scale the solution into other business operations.
The hype train is slowing down – executives are growing impatient and want AI payback
In the early days of AI, when business executives were investing heavily, a pragmatic approach would have been perceived as unambitious. But today, with many AI projects failing to meet expectations, impatience is growing. Investments have been made, platforms have been brought, resources have been onboarded, and partnerships have been formed. Key stakeholders want to see concrete results and realise that a new approach is required. It’s about many smaller wins, building internal competence, choosing partners, putting the frameworks in place, and then accelerating.
Leaders must let go! Time to prioritise speed and agility over centralised control
One of the most dangerous things leaders can do is identify a great initiative but fail to support it properly because their resources are finite. The role of a strong leader in an AI integration project is to focus resources where they matter for the bottom line.
Business leaders must realise that they cannot be an expert in every field, and a philosophy of delegation of authority needs to be established. Bottlenecks are often caused in AI integration when everything must go through leadership teams. They’re not the most skilled decision-makers in every domain, so decision-making should be closest to the customer or closest to the technology experts. From here, leaders cannot fall into the trap of micromanaging as another bottleneck will form.
Put your trust in the team of experts
Leaders need to trust their workers, arm them with the tools they need, provide them with a strong governance framework, especially around security, and then relinquish some control and monitor progress. Through doing this, teams can work without constraints, and leaders can identify when teams are making progress and give them what they need to scale it.
Leaders must recognise that trust is two-way
As with any new technology that businesses look to integrate to drive value, leaders are often faced with caution from their workforce, and this has been the case for over a third of UK small businesses. However, there are three strategic shifts that leaders need to focus on in response to their workforce when leveraging AI:
- A turbulent but temporary transition: One of the major concerns with AI integration is the pressure it is putting on white-collar and knowledge workers. While there will be a turbulent period as this happens, leaders need to look at it from the perspective that AI is making better use of workers’ time. The UK is facing a major labour shortage that could cost £30 billion a year if not addressed, and AI has the potential to bridge this gap. AI can help workers become more productive by completing monotonous jobs quicker, which gives workers more time to focus on value-added tasks, and in the long-term, workers will move on to new, more meaningful roles.
- Reskilling and upskilling: The workforce is an integral part of any business, and right now, they need to evolve to work in conjunction with AI. Leaders need to provide their workers with the best possible opportunities, as more job categories get created, they must reskill and upskill workers to ensure their workforce transitions seamlessly.
- Striking the balance between human and AI: When technology is in the right hands, it creates an undeniable performance gap, just look at the tractor or conveyor belt and the productivity boosts they brought. Successful AI integration doesn’t hinge on leaders choosing AI over humans but finding the right balance. Identifying people with the right focus and ability to adapt, and moving the workforce into roles where they complement AI rather than compete with it.
AI is here to stay – over to leaders to lead in the AI era
The surge in failed AI projects in recent years must act as a reminder to businesses that hype is not a replacement for strategy. Leadership teams must take a pragmatic approach. One that focuses on bridging the gap between technological development and human ability by setting executive-level expectations, shifting from centralised control to a philosophy of delegation to break human bottlenecks, and homes in on the speed of scaling smaller wins.
When AI is integrated in conjunction with human intelligence, rather than a replacement, leadership will unlock more than just productivity and efficiency benefits.
