Many organisations have already invested in AI, yet fewer are seeing the measurable business impact they expected. In our recent webinar, we explored why that gap often emerges and what organisations can do to close it.
During the session, we received more questions than we had time to answer. To continue the conversation, we asked our People and Organisation Lead, Ana Anaya, to share her perspective on some of the most common challenges organisations face in turning AI experimentation into sustained value.
Below, Ana answers questions submitted during the webinar, covering topics such as AI adoption, governance, scaling successful use cases, and where organisations may be overestimating the technology. These are issues we see regularly in our work with clients as we help them move from AI investment to measurable impact.
How do you define “AI value” at board level?
At board level, AI value should be defined in terms of measurable business outcomes, not tool usage. That means improvements in areas such as revenue growth, customer experience, operational efficiency, or risk management. Many organisations see AI activity increasing, but struggle to connect that activity to meaningful outcomes. Board-level value comes when AI is embedded in key business processes and aligned with strategic priorities, so its impact can be measured against metrics leadership already cares about.
At what point should organisations stop experimenting and start scaling AI initiatives?
Experimentation is essential early on, but organisations should start thinking about scaling once a use case clearly demonstrates repeatable value. The key signal is when an AI application improves a real business process and can be applied consistently across teams or functions. At that point, the focus should shift from individual productivity gains to organisational impact, ensuring the use case is supported by appropriate governance, an adoption approach, and integration into existing workflows.
What typically blocks AI adoption: leadership hesitation or frontline resistance?
In most cases, the blocker is neither outright leadership hesitation nor frontline resistance, but a gap in structured adoption. Leaders often invest in tools but underestimate the effort required to embed them into daily work. At the same time, frontline teams may not fully understand how AI applies to their specific roles. Successful organisations address this by treating AI adoption as a change programme, with targeted training, role-based use cases, and clear leadership sponsorship.
We are seeing lots of isolated (but powerful) use cases across the organisation. What steps would you suggest to rationalise and share best practices?
This is a common stage in AI maturity. Organisations should first catalogue existing use cases and identify which are delivering measurable value. From there, the focus should shift to creating mechanisms to scale successful examples across the organisation. This may involve establishing a central AI governance or enablement group, encouraging cross-team knowledge sharing, and embedding successful use cases into standard processes rather than leaving them as isolated experiments.
Where are organisations overestimating AI’s capabilities today?
Many organisations overestimate how quickly AI tools alone can deliver enterprise-wide impact. While the technology is evolving rapidly, value often depends on how well AI is integrated into processes, supported by training, and aligned to business priorities. Without that foundation, even powerful tools can produce limited results. In practice, the biggest challenge is rarely the technology's capability, but the organisational work required to embed it effectively.
What typically prevents AI initiatives from moving forward?
The most common blocker is an inability to clearly demonstrate value. If organisations can’t quantify impact or explain why an AI initiative matters, momentum quickly fades and leadership support weakens. AI projects often stall when they are launched without a clearly defined business problem or measurable outcome. The most effective way to overcome this is to start with a tangible challenge and define success upfront. Clarity of purpose creates buy-in, alignment, and sustained progress.
You mentioned seeing AI as a partner, not a replacement. Are there any practical steps you can suggest to help shift employee mindsets from fear of AI to confident, responsible use?
Shifting the mindset starts with how organisations introduce AI to their people. If AI is framed primarily as a cost-saving tool, employees naturally assume it threatens their roles. Instead, leaders should position AI as a capability that augments human work by removing repetitive tasks and enabling employees to focus on higher-value activities.
Practically, this means providing role-specific examples of how AI supports day-to-day work, investing in targeted training, and encouraging safe experimentation within clear governance boundaries. When people see AI improving how they do their job, confidence quickly replaces fear.
How can organisations build AI confidence without encouraging misuse or shadow AI?
Building confidence starts with clear guidance on which tools are approved and how to use them. When expectations are explicit, the risk of shadow AI reduces significantly. Beyond guardrails, organisations must invest in structured, role-based training that supports different user types from innovators to occasional users. Confidence grows when people feel equipped to experiment safely within defined boundaries. Combining clear governance with targeted enablement creates responsible adoption rather than uncontrolled experimentation.
How do you secure leadership buy-in when AI is seen as just an IT initiative?
Leadership buy-in improves when AI is framed in business terms rather than technical ones. Rather than focusing on tool deployment, connect AI use cases directly to outcomes that matter to that leader, such as revenue growth, customer impact, or operational efficiency. Demonstrating how AI solves a specific problem they care about is far more compelling than presenting abstract capability. Co-creating small, relevant use cases can build credibility and momentum, turning scepticism into sponsorship over time.
We’ve rolled out an AI tool, but adoption is patchy. What should we do?
Patchy adoption often signals that enablement has been too generic or passive. Announcements and static training materials rarely drive behavioural change. Start by reassessing how you introduced the tool and shift toward interactive, team-based experimentation. Encourage managers to embed AI into regular team rituals and use real examples to demonstrate value. Once users see practical benefits in their day-to-day work, adoption accelerates. Treat rollout as an ongoing change programme, not a one-off communication exercise.
What if an organisation has already invested in AI but isn’t seeing the value it expected?
This is a situation we see quite often. Many organisations have deployed AI tools and encouraged experimentation, but the impact isn’t always translating into meaningful business outcomes.
In most cases, the challenge isn’t the technology itself. Value can stall due to barriers to adoption, unclear ownership of AI initiatives, weak integration into business processes, or a lack of governance and guardrails.
To help organisations address this, we developed AI Realise. It’s a practical framework designed to help organisations identify what may be preventing AI value from being realised and provide a structured approach to overcoming those blockers. The goal is simple: move from AI experimentation to measurable, sustainable business impact. You can find out more information about AI Realise here.
If you’d like to explore the topic in more detail, you can still watch the full webinar on demand here.
And if you have a question we haven’t covered, feel free to reach out to the team here.
.jpg)